Skip to main content
NCBI home page
Wellcome Open Research logoLink to Wellcome Open Research
. 2025 May 13;10:112. Originally published 2025 Feb 28. [Version 2] doi: 10.12688/wellcomeopenres.23509.2

EFFICIENT RIBOSOMAL RNA DEPLETION FROM DROSOPHILA TOTAL RNA FOR NEXT-GENERATION SEQUENCING APPLICATIONS

Omkar Koppaka 1, Shweta Tandon 1, Ankita Chodankar 1,2, Awadhesh Pandit 2, Baskar Bakthavachalu 1,a
PMCID: PMC12123297  PMID: 40454261

Version Changes

Revised. Amendments from Version 1

In response to reviewers’ suggestions, we have made revisions to improve the clarity and completeness of our manuscript. We have included siTOOLs, a kit specifically developed for Drosophila, in both the introduction and Table 1. We have edited the methods and results sections to include the percentage of rRNA covered by the probes. We have also clarified the polyA enrichment and rRNA depletion steps in both the methods and discussion sections. Importantly, we identified and corrected an error in the previous version of Supplementary Table 2, where only single-end sequencing depth was reported for the rRNA depletion (kit-based) method. Furthermore, we have revised the table to present sequencing depth and mapping percentages for individual replicates.

Abstract

We developed a cost-effective enzyme-based rRNA-depletion method tailored for Drosophila melanogaster, addressing the limitations of existing commercial kits and the lack of peer-reviewed alternatives. Our method employs single-stranded DNA probes complementary to Drosophila rRNA, forming DNA-RNA hybrids. These hybrids are then degraded using the RNase H enzyme, effectively removing rRNA and enriching all non-ribosomal RNAs, including mRNA, lncRNA and small RNA. When compared to a commercial rRNA removal kit, our approach demonstrated superior rRNA removal efficiency and mapping percentage, confirming its effectiveness. Additionally, our method successfully enriched the non-coding transcriptome, making it a valuable tool for studying ncRNA in Drosophila. The probe sequences and rRNA-depletion protocol are made freely available, offering a reliable alternative for rRNA-depletion experiments.

Keywords: Drosophila, rRNA, RNase H, Next-Generation Sequencing

Plain language summary

Drosophila melanogaster is a model organism that plays an important role in understanding various conditions in health and disease. RNA helps in the maintenance of organismal health, while dysregulation leads to diseases. Various methods are available to study RNA expression at small- and large-scale levels. One of the most useful of these is the method called RNA sequencing. However, for good quality data, the rRNA, which is around 80% of total RNA, has to be removed. We use custom single-stranded DNA probes, which bind to the rRNA molecules. These rRNA:ssDNA complexes can then be removed using RNAse H, which specifically depletes RNA:DNA hybrids. Using this method, we were able to deplete ~97% of rRNA.

Introduction

Drosophila melanogaster has emerged as a key model organism due to rapid generation time, easy maintenance, well-characterized genome and low experimental cost ( Mirzoyan et al., 2019; Tolwinski, 2017). With a genome that shares ~60% homology with humans of which 75% are linked to human diseases ( Adams et al., 2000; Banfi et al., 1996; Fortini et al., 2000), Drosophila plays a critical role in studying numerous conditions such as neurodegeneration, cardiac diseases, metabolic disorders and even rare human diseases ( Goodman & Bellen, 2022; Kim et al., 2021; Link & Bellen, 2020; Souidi & Jagla, 2021). Beyond disease models, Drosophila is also essential for exploring normal physiology and gene regulation, as its transcriptome provides insights into conserved cellular mechanisms and gene interactions fundamental to development and cellular function ( Hardin et al., 1990; Harvey et al., 2003; Nüsslein-Volhard & Wieschaus, 1980). This model organism enables researchers to investigate gene regulation under a range of conditions, advancing our understanding of both healthy and pathological states.

RNA expression plays a crucial role in maintenance of physiological balance, as evidenced by the fact that dysregulation of RNA leads to several pathological conditions ( Bradley & Anczuków, 2023; Nemeth et al., 2024). To understand the roles of RNA expression in physiological and pathological conditions, identifying both global and tissue-specific gene expression patterns is crucial. Earlier methods, such as Sanger sequencing, Northern blotting, and qRT-PCR, were both labor and time-intensive and could only analyze a small portion of the transcriptome. The introduction of high-throughput methods, such as RNA sequencing and its variants, have transformed transcriptomic research ( Daines et al., 2011; Tang et al., 2009; Wang et al., 2009). RNA sequencing now allows for comprehensive profiling of gene expression, providing a broad and detailed view of transcriptomic dynamics. These advancements have significantly improved our capacity to study complex biological processes and diseases at a molecular level, advancing our understanding of various complex conditions.

Ribosomal RNA (rRNA), the most abundant type of RNA, typically constitutes about 80% of total RNA ( Singer & Berg, 1991), posing a significant challenge for profiling the RNA types relevant for gene expression studies. Therefore, it must be removed from total RNA before sequencing, to allow for the enrichment of high-value targets ( Huang et al., 2011; O’Neil et al., 2013). Currently, two methods exist for rRNA removal: polyA-enrichment and rRNA-depletion. PolyA-enrichment is widely used because of its cost-effectiveness and ability to work effectively with low sequencing depth, often providing better coverage compared to rRNA depletion at equivalent sequencing depth. Unfortunately, polyA-enrichment sometimes leads to a 3’ end bias ( Tariq et al., 2011) which renders it less effective for low-quality RNA or Formalin-Fixed Paraffin-Embedded (FFPE) tissue samples, which generally have degraded RNA ( Kellman et al., 2021; Zhao et al., 2014). While this means that rRNA that lack polyA tail are eliminated, it also leads to the loss of many non-coding RNA (ncRNA) lacking a polyA tail. rRNA-depletion, which focuses on the elimination of rRNA rather than the selection of a particular class of RNA serves as a much better alternative for low-quality and degraded RNA while also enriching the non-coding RNA that would otherwise be lost in the polyA-selection method ( Cui et al., 2010; Kissopoulou et al., 2013). Depletion of rRNA generally depends on the principle of hybridization of single-stranded rDNA probes to rRNA. Thus, for efficient rRNA depletion and minimal off-target activity, these probes need to be highly specific for the species in which rRNA depletion must be done. This restricts the use of the kits to the intended organisms and those that share their conserved rRNA regions ( Kraus et al., 2019). rRNA sequences are relatively conserved across eukaryotes, especially higher eukaryotes. Hence, kits developed for humans also work for rats and mice. However, in the case of insects, the ribosome biogenesis pathways are different. The 28S rDNA has 2 types of insertions; while Type I insertion resembles mammalian rDNA with a single block for 28S rDNA, Type II insertion has 2 blocks of 28S rRNA that are separated from each other by 5.4 kb of DNA ( Glover & Hogness, 1977; Pellegrini et al., 1977; Tautz et al., 1988; White & Hogness, 1977). 28S rRNA is transcribed from these rDNA along with 5.8S and 18S as a single large precursor. Following splicing, the 28S rRNA undergoes further processing through an unknown mechanism which leads to fragmentation of the 28S rRNA into α and ß fragments ( Dawid & Wellauer, 1978; Kidd & Glover, 1981; Winnebeck et al., 2010). Due to this fragmentation of the insect rRNA during processing, the commonly available kits used for vertebrate rRNA depletion, such as RiboZero and RiboZero Gold do not completely remove 5S rRNA or 28S rRNA, which leads to low mapping and sequence coverage. Thus, most commercially available kits do not recommend using kits developed for a particular organism or group of organisms to be used for rRNA depletion in other organisms ( Figure 1A). This inefficiency can negatively impact downstream analyses, such as sequence mapping and coverage. Due to the lack of an efficient protocol, several previous studies have performed rRNA depletion using commercial kits designed for vertebrates in Drosophila ( Table 1).

Figure 1. rRNA biogenesis and depletion methods.

Figure 1.

Open in a new tab

A) The rRNA biogenesis pathways differ in Drosophila and Mammals. While mammals have one fragment of 28S rRNA, Drosophila 28S rRNA has 2 fragments: α and ß. ( B) A schematic for using biotinylated probes or RNAse H for rRNA-depletion.

Table 1. rRNA-depletion kits previously used for Drosophila samples.

Name of Kit Recommended
Species		 Notes Reference
Tecan Ovation ® RNA-Seq System V2 - cDNA synthesis without polyA
selection or rRNA depletion		 ( Hughes et al., 2012)
Zymo-Seq RiboFree Total RNA library Kit Any 20% of Insect rRNA remains
undepleted		 ( Talross & Carlson, 2023)
Illumina Ribo-Zero™ Gold Kit HMR Currently Unavailable ( Chen et al., 2016)
Illumina Ribo-Zero rRNA removal kit HMR Currently Unavailable ( Pritykin et al., 2017)
Thermo Fisher Scientific RiboMinus™
Eukaryote Kit for RNA-Seq		 HMR Low compatibility with Drosophila
as stated by the manufacturer		 ( Lefebvre et al., 2017; Lim et al., 2013; Mugat et al., 2015)
QIAseq FastSelect-rRNA Fly kit Fly Inhibits rRNA reverse transcription to cDNA ( de Queiroz et al., 2025; Malwade et al., 2024; Ye et al., 2024)
riboPOOL kit Fly Available for both intact and degraded RNA ( Kaneko et al., 2024; van Lopik et al., 2023)
Open in a new tab

(HMR – Human, Mouse and Rat).

However, due to these reasons mentioned earlier, kits developed for mammals that were used in earlier studies are no longer available or not recommended for use with Drosophila samples. However, there are a few promising developments in Drosophila-specific rRNA depletion kits like Zymo's Seq RiboFree Total RNA Library Kit, which claims to remove approximately 80% of Drosophila rRNA. Also, Qiagen has developed a QIAseq FastSelect–rRNA Fly Kit specifically for Drosophila rRNA removal. Similarly, siTOOLs has developed Fly specific riboPOOL kits which use biotinylated probes to deplete rRNA in intact and fragmented RNA samples. Despite the availability and time-saving benefits of commercial kits, they are often inefficient and exorbitantly priced, leading some labs to opt for in-house preparation of rRNA-depletion methods.

There are two primary methods for removing rRNA, each with its own set of advantages and limitations. The first approach, known as pulldown, involves the use of biotinylated DNA oligos that are complementary to rRNA sequences. These oligos hybridize with the rRNA, and the resulting RNA-DNA hybrids are captured using streptavidin-conjugated beads. Once bound, the rRNA can be efficiently removed ( Bhagwat et al., 2014; Chen & Duan, 2011; Culviner et al., 2020). While this method is highly specific and effective for rRNA depletion, it is less suitable for fragmented or degraded RNA samples. The high specificity of the pulldown probes means that if the RNA is fragmented, only portions of the rRNA may bind to the probes, leaving residual rRNA in the sample. Additionally, it requires a large initial amount of starting material, which may not always be feasible. An alternative method is enzymatic depletion, which also relies on the formation of RNA-DNA hybrids ( Baldwin et al., 2021). In this case, the RNase H enzyme is employed to degrade the rRNA specifically bound to the complementary DNA oligos ( Figure 1B). This method has the advantage of being effective even with low-quality or degraded RNA ( Wahl et al., 2022). However, different species require the design of species-specific probes for optimal performance. Both techniques have been used to develop in-house rRNA-depletion kits for Drosophila. Biotinylated probes have been developed to remove various rRNA fractions ranging from 2S to 28S rRNA ( Fowler et al., 2018; Thompson et al., 2020). RNase H dependent enzymatic degradation has also been used for the depletion of 5.8S, 18S and 28S rRNA ( Haugen et al., 2024). However, the probe sequences have not been made publicly available.

To address these challenges, we have designed rRNA probes specific for Drosophila melanogaster rRNA sequences and demonstrate effective and reproducible depletion of rRNA from adult Drosophila brain samples. We make the probe sequences freely available for use by the research community. Our study provides a comprehensive protocol for rRNA-depletion using the RNase H method, from probe design to data analysis, enabling other researchers to replicate and adapt this approach for their studies in Drosophila melanogaster and other closely related species.

Methods

Fly rearing and genetics

Drosophila Canton-S stocks were maintained at 25 ±1°C on cornmeal-sugar-agar media.

RNA extraction

RNA was extracted from the brains of 5-day-old adult Drosophila (N=15) dissected in 1X PBS using TRIzol (Cat No. 15596026, Thermo Fisher Scientific). Briefly, ~15 brains were homogenized in 500 µl of TRIzol until fully dissolved. Subsequently, 100 µl of chloroform was added, and the mixture was vortexed briefly and centrifuged at 13,000g for 10 minutes at 4°C. The aqueous phase was transferred to a fresh tube containing 250 µl of isopropanol to precipitate RNA. The RNA pellet was washed with 70% ethanol and dissolved in 10 µl of nuclease-free water. RNA quality was evaluated using the Agilent Tapestation 4200 before cDNA synthesis, library preparation, and sequencing.

Probe design and reconstitution

Sequences for 5S (URS00003B4856_7227), 5.8S (URS00005FF212_7227), 18S (URS0000A575BB_7227) and 28S (URS0000A53741_7227) rRNA were sourced from RNAcentral ( Sweeney et al., 2019). Probes were designed using the NEBNext Custom RNA Depletion Design Tool. These probes were analyzed using BLAST to assess potential off-target effects, employing default parameters with specific adjustments for "Organism" ( Drosophila melanogaster) and "Optimize for" (blastn) ( Altschul et al., 1990). Subsequently, a final list of probes which covered 98% of rDNA (Table S1) was synthesized by Integrated DNA Technologies, Pvt. Ltd. and reconstituted in nuclease-free water to achieve a stock concentration of 250 µM. These probes were then pooled together, resulting in a final concentration of 2 µM per probe.

Depletion of rRNA using custom probes

Total RNA which was extracted was split for PolyA enrichment and rRNA depletion. All libraries were prepared using 100 ng of total RNA for poly(A)-selection method and 1 µg of total RNA for rRNA depletion. 2 biological replicates were prepared for PolyA enrichment and rRNA depletion using custom probes. For Poly(A)-enrichment of mRNA, Illumina libraries were prepared using NEBNext Ultra II Directional RNA Library Prep kit (E7765L) and sequenced with Illumina NovaSeq 6000 system. rRNA-depletion was performed as recommended by NEBNext RNA Depletion Core Reagent protocol. Briefly, the RNA integrity was confirmed on an Agilent Tapestation 4200 and samples with at least 7.0 RIN values were used for experiments. The total RNA (500–1000 ng), 2 μL of the 2μM probe pool, 2 μL of the 2μM NEBNext probe hybridization buffer (E6314) and nuclease-free water were mixed to a total volume of 15 μL. This mixture was then incubated at 95°C for 2 min, a touchdown from 95°C to 22°C (0.1°C/sec) and a 5-min hold at 22°C. NEBNext RNA Depletion Core Reagent Set (E7865S) was used for depletion of rRNA where we set up a reaction with 2 μL NEBNext RNase H Reaction Buffer, 2 μL NEBNext Thermostable RNase H and 1 μL nuclease-free water and incubated for 30 min at 50°C. The RNase H treated sample was digested with DNase I and purified using NEBNext RNA Sample Purification Beads. The concentrations of purified RNA samples were measured using the Qubit HS RNA Assay. The purified rRNA-depleted samples were taken for library preparation using NEBNext Ultra™ II Directional RNA Library Prep with Sample Purification Beads (Catalogue no-E7765L) as per the manufacturer's protocol. The libraries were sequenced on the Illumina NovaSeq 6000 platform with a 2x100bp read length. Similarly, rRNA-depletion was performed as recommended by QIAseq FastSelect – rRNA Fly Kits protocol (Cat. No. / ID: 333262) using total RNA extracted from Drosophila S2 cells. The rRNA-depleted samples were processed using a similar protocol as described above.

Data analysis

Sequenced files in FASTQ format were initially assessed for quality using FastQC ( Andrews, 2017/2024). Trimmomatic was used to trim the reads as per the following parameters: -phred33 HEADCROP:6 LEADING:25 TRAILING:25 AVGQUAL:25 MINLEN:19 ( Bolger et al., 2014). Once high-quality reads were confirmed, alignment was performed against the Drosophila dm6 reference genome, using STAR ( Dobin et al., 2013). The count matrix was generated using the featureCounts package ( Liao et al., 2014). Differential expression analysis was conducted using the DESeq2 package in RStudio ( Love et al., 2014). Visualization of results was achieved with EnhancedVolcano for volcano plots and pheatmap for heatmaps ( Kolde, 2010). Picard ( Broadinstitute/Picard, 2014/2024) was utilized to calculate gene region coverage, while custom scripts were employed for generating Venn diagrams and bar plots. Ribodetector was used to estimate the rRNA content in each sample ( Deng et al., 2022, p. 2). The softwares used in the study are listed in Table S4.

Results

Design and selection of custom probes for Drosophila rRNA-depletion

The effectiveness of rRNA removal using RNase H relies on successful hybridization between rRNA and single-stranded DNA probes with sequence homology. Poorly designed probes can impact downstream steps and compromise data quality, making probe design a critical step in rRNA depletion. For this purpose, we downloaded the sequences of Drosophila melanogaster 5S (URS00003B4856_7227), 5.8S (URS00005FF212_7227), 18S (URS0000A575BB_7227) and 28S (URS0000A53741_7227) rRNA from RNACentral and designed specific probes using the NEBNext Custom RNA Depletion Design Tool. This process generated 2, 2, 34, and 64 probes for 5S, 5.8S, 18S, and 28S rRNA, respectively. Since RNase H degrades RNA-DNA hybrid regardless of sequence, non-specific probes could inadvertently deplete sequences within our transcriptome of interest. To prevent this, we performed sequence homology searches using BLAST for all the probe sequences, ensuring that non-target RNA with homology to our probes would not be removed unintentionally.

Using the NEBNext Custom RNA Depletion Design Tool, our initial design generated 164 probes, many of which had overlapping sequences. After manual curation of the overlapping probes, we finalized a set of 102 probes (Table S1) that comprehensively covered 98 % of the 5S, 5.8S, 18S, and 28S rRNAs with minimal gaps.

Custom-generated probes for RNase H based rRNA-depletion performed better than the commercial kit

RNA was extracted from Drosophila brains, and its quality was evaluated using an Agilent Tapestation 4200. The freshly isolated total RNA showed a prominent rRNA peak corresponding to the 18S size ( Figure 2B-i). Unlike mammalian rRNA, which produces two distinct peaks at approximately 1.9 kb and 5 kb for 18S and 28S respectively, Drosophila 28S rRNA is naturally fragmented into 28 S α and 28S ß which migrates close to the 18S rRNA, resulting in two prominent peaks at approximately 2 kb ( Winnebeck et al., 2010). This total RNA was used for rRNA-depletion using the synthesized custom rRNA probes and RNase H as described in the methods section. The rRNA-depleted samples were evaluated using an Agilent Tapestation 4200, which showed a complete absence of peaks corresponding to 18S and 28S rRNA bands post-depletion ( Figure 2B-ii). An equivalent quality of total RNA extracted from S2 cells was used to deplete rRNA using the QIAseq FastSelect –rRNA Fly Kit. In parallel, we also enriched polyA RNA from both samples to compare the levels of rRNA and ncRNA with respect to rRNA-depleted samples.

Figure 2. Efficient depletion of rRNA with RNase H using custom-generated probes.

Figure 2.

Open in a new tab

( A) Schematic workflow for probe design and rRNA depletion using RNAse H. ( B) Tape station analysis shows the reduction in 18S and 28S rRNA after depletion. ( C) Mapping percentages of samples sequenced using commercial kit (Qiaseq fast Select-fly rRNA depletion kit) and custom probes (RNAse H Method). ( D) Mapped rRNA read percentage from the custom probe and RNAse H method and Qiaseq Fastselect - fly rRNA depletion kit.

Next-generation sequencing (NGS) libraries were generated for rRNA-depleted and polyA-selected RNA using NEBNext Ultra™ II Directional RNA Library Prep Kit. The libraries were sequenced on Illumina NovaSeq 6000, and the reads and quality scores are tabulated in Table S2. FastQC was used to assess the quality of sequenced reads that consistently returned a high quality across all samples, with per-base sequence quality scores exceeding 35.

With the data quality confirmed, the sequencing reads were trimmed as described in the methods section, before mapping to the Drosophila melanogaster reference genome (dm6) using STAR. In the samples where custom probes were used for rRNA depletion, the mapping was about 83.47%, which was quite comparable to the polyA-enriched sample. However, the commercial kit-based rRNA depletion removed the rRNA inefficiently, leading to just ~ 40% mapping as compared to the 83.2% mapping in the polyA-enriched sample ( Figure 2C, Table S2). The effectiveness of rRNA removal was determined using Ribodetector, which showed that samples prepared with the QIAseq FastSelect –rRNA Fly kit retained ~39% rRNA, which explains the poor mapping rate ( Figure 2D-i). On the other hand, the custom probe-based method retained just 3.5%, showing efficient rRNA depletion ( Figure 2D-ii). Given the lack of a sufficient number of publications using the QIAseq FastSelect -rRNA Fly kit, it might need optimization. However, our findings indicate that, in its current iteration, the probe-based RNase H method provides superior performance for rRNA depletion.

Non-coding RNAs were identified in the rRNA-depleted samples

The purpose of using rRNA depletion sequencing rather than polyA-selection is to capture a broader population of ncRNAs, which typically lack polyA tail. To evaluate whether the rRNA-depletion method successfully enriched the non-coding transcriptome, we generated count matrices for both rRNA-depleted and polyA-enriched datasets from BAM files using featureCounts. We then performed differential expression analysis by comparing the rRNA count matrix to the polyA count matrix. A volcano plot illustrated the enrichment of RNAs in the rRNA-depleted samples, with several ncRNAs such as CR434459, snRNA:7SK, and hsr-omega prominently upregulated in rRNA-depleted samples ( Figure 3A). A list of these upregulated genes, including their log2 fold changes and adjusted p-values, can be found in Table S3. A heatmap to display the differences in enriched genes between the rRNA-depleted and polyA-selection samples is shown in Figure 3B.

Figure 3. Non-coding RNAs were enriched in the rRNA-depleted samples.

Figure 3.

Open in a new tab

( A) A Volcano plot showing RNA targets enriched in rRNA-depleted samples as compared to the polyA-enriched samples. ( B) Heat map of > 2-fold change enriched genes with < 0.05 padj value (Fold change values in Table S3). ( C) rRNA-depletion increased intronic coverage as compared to the polyA-enriched samples.

Furthermore, we classified the top 15 ncRNA, ranking them by log2 fold change while excluding those with adjusted p-values greater than 0.05. Most of the identified ncRNA were long non-coding RNAs (lncRNAs), such as CR44157 and CR44474, followed by small nucleolar RNAs (snoRNAs) ( Table 2).

Table 2. Top 15 non-coding RNA.

Name Classification
snoRNA: CG32479-b sno RNA
snoRNA: Or-aca4 sno RNA
mir-4969 miRNA
mir-2279 miRNA
CR44157 lncRNA
mir-317 miRNA
CR44474 lncRNA
CR43399 lncRNA
CR45256 lncRNA
snoRNA: Psi28S-2622 sno RNA
CR44279 lncRNA
mir-4982 miRNA
CR44496 lncRNA
CR44034 lncRNA
CR44070 lncRNA
Open in a new tab

In addition to the detection of mRNA and ncRNA, rRNA depletion can capture nascent transcripts, and thus, there is a possibility of higher intronic sequences in rRNA-depletion as compared to polyA-enrichment samples ( Zhao et al., 2018).To confirm this, we used Picard to assess coverage across different genomic regions in both rRNA depleted and polyA samples. Intronic sequences were notably more abundant in the rRNA depleted samples compared to the polyA samples ( Figure 3C). This observation suggests that, while not highly efficient, the rRNA depletion method can detect nascent mRNAs.

Given this, rRNA depletion can be considered a suitable approach for most studies, particularly considering the increasing recognition of the functional importance of ncRNAs. Unless the study specifically requires the detection of low-abundance mRNA targets, rRNA depletion offers a compelling alternative, enabling the exploration of both coding and non-coding RNA dynamics.

Discussion

Despite being a vital model organism, Drosophila has been largely overlooked when it comes to the development of rRNA-depletion kits. Most commercially available kits are specifically optimised for mammalian systems, primarily humans, mice, and rats (HMR). This bias has forced researchers working with Drosophila to adapt HMR-optimised kits for rRNA depletion in their studies. However, this approach is suboptimal due to significant differences between Drosophila and HMR rRNA structures, particularly the fragmented nature of the 28S rRNA in Drosophila compared to the single, continuous 28S rRNA found in mammals. These differences affect the efficiency of rRNA depletion in Drosophila, highlighting the need for a species-specific solution.

To fill this gap, we developed a Drosophila-specific rRNA depletion method. Although some laboratories have created rRNA depletion techniques for Drosophila, their complete probe sequences are unavailable, and their efficiency has not been thoroughly validated or compared to commercial kits ( Haugen et al., 2024; Thompson et al., 2020). Existing commercial kits often rely on biotinylated probes, which are inefficient for fragmented or low-quality RNA ( Culviner et al., 2020). To address these challenges, we designed custom probes specifically targeting Drosophila rRNA and paired them with RNase H for efficient depletion. This approach is optimized to enhance efficiency and produces rRNA-depleted RNA suitable for diverse downstream applications, including qRT-PCR, microarray analysis, and NGS.

The NEB Probe Designer tool, used for designing rRNA probes, generates multiple probes with overlapping sequences. To minimize costs and avoid redundancy, these overlapping probes were manually reviewed and eliminated. Due to the short sequence of 2S rRNA (~30 nucleotides), the tool did not design a probe for it. All libraries were prepared using 100 ng of total RNA for poly(A)-selection method and 1 µg of total RNA for rRNA depletion. We did not observe 2S rRNA in our sequencing reads, as it might have been depleted in the RNA purification step following enzymatic digestions. Importantly, the custom probe method produced reproducible results across both replicates.

Further, this method is best suited for the identification of ncRNA in fly tissues. Differential analysis identified several ncRNAs, particularly lncRNAs, were significantly enriched in the rRNA-depleted samples compared to the polyA-enriched ones ( Figure 3A). Previous studies have reported that nascent mRNAs, which typically lack a polyA tail, are not efficiently detected using polyA-enriched samples ( Yang et al., 2011; Zhao et al., 2018). However, rRNA depletion allows nascent RNA detection from the intronic reads ( Figure 3C).

Interestingly, we also detected a few microRNAs (miRNAs) ( Table 2). Given their small size (20–22 nucleotides), miRNAs are typically not expected in rRNA-depleted samples and require specialized sequencing protocols that include specific size-selection steps for enrichment. We hypothesize that these detected miRNAs may belong to a subset of highly abundant ones.

Despite these strengths, a few limitations of the method should be noted. Samples sequenced following rRNA depletion and mapping to the reference genome had a small percentage of reads that were too short to align effectively ("% of reads unmapped: too short"). These could be due to leftover rRNA fragments from depletion. Further STAR documentation does note that “poor sequencing quality, over-trimming, or contamination with ribosomal RNA fragments can yield reads classified as too short to map”. Despite this, we obtained mapping percentages of custom probes mediated rRNA depleted samples that were comparable to the polyA selection method ( Figure 2C).

These findings underscore the suitability of rRNA depletion for capturing both coding and non-coding RNA populations, including nascent RNA. Given the high-quality results observed, this method is a robust alternative to commercial kits, and we are confident that this custom probe-based rRNA depletion method will be used more often while sequencing Drosophila species as well as closely related insect transcriptomes.

Ethical and consent statement

The experiments presented in the study were carried out at an appropriate biosafety level laboratory with institute biosafety approvals.

Acknowledgments

We thank members of the Bakthavachalu lab and Amanjot Singh, for their useful comments on the manuscript. The fly facility at Bangalore Life Science Cluster (BLiSC) provided support with fly stock supply as well as generation of transgenics and next-generation genomics facility at BLiSC provided NGS service. Finally, we would like to thank the PARAM Himalaya cluster at IIT Mandi for providing the computational resources and technical support for carrying out various analysis.

Funding Statement

This work was supported by the Wellcome Trust DBT India Alliance [IA/I/19/1/504286]; and DBT-RA fellowship [DBT-RA/2023/July/N/4009; to ST]. The funders had no role in study design, data collection and interpretation or the decision to submit the work for publication.

The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

[version 2; peer review: 2 approved]

Data availability

Underlying data

NCBI Gene Expression Omnibus: RNA Sequencing data from NCBI GEO. Accession ID: GSE282990; http://identifiers.org/geo:GSE282990 ( Koppaka et al., 2024a).

Extended data

The extended data from this study can be accessed from:

Figshare: Efficient ribosomal RNA depletion from Drosophila total RNA for next generation sequencing applications. https://doi.org/10.6084/m9.figshare.28093562 ( Koppaka et al., 2024b).

The project contains the following extended data:

  • -

    Table S1: List of custom probes used for RNase H mediated rRNA-depletion

  • -

    Table S2: NGS data mapping details

  • -

    Table S3: Differential Expression analysis results

  • -

    Table S4: List of software used in the study

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

References

  1. Adams MD, Celniker SE, Holt RA, et al. : The genome sequence of Drosophila melanogaster. Science. 2000;287(5461):2185–2195. 10.1126/science.287.5461.2185 [DOI] [PubMed] [Google Scholar]
  2. Altschul SF, Gish W, Miller W, et al. : Basic Local Alignment Search Tool. J Mol Biol. 1990;215(3):403–410. 10.1016/S0022-2836(05)80360-2 [DOI] [PubMed] [Google Scholar]
  3. Andrews S: S-andrews/FastQC.[Java], (Original work published 2017),2024. Reference Source [Google Scholar]
  4. Baldwin A, Morris AR, Mukherjee N: An easy, cost-effective, and scalable method to deplete human ribosomal RNA for RNA-seq. Curr Protoc. 2021;1(6): e176. 10.1002/cpz1.176 [DOI] [PubMed] [Google Scholar]
  5. Banfi S, Borsani G, Rossi E, et al. : Identification and mapping of human cDNAs homologous to Drosophila mutant genes through EST database searching. Nat Genet. 1996;13(2):167–174. 10.1038/ng0696-167 [DOI] [PubMed] [Google Scholar]
  6. Bhagwat AA, Ying ZI, Smith A: Evaluation of ribosomal RNA removal protocols for Salmonella RNA-Seq projects. Adv Microbiol. 2014;4(1):25–32. 10.4236/aim.2014.41006 [DOI] [Google Scholar]
  7. Bolger AM, Lohse M, Usadel B: Trimmomatic: a flexible trimmer for illumina sequence data. Bioinformatics. 2014;30(15):2114–2120. 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bradley RK, Anczuków O: RNA splicing dysregulation and the hallmarks of cancer. Nat Rev Cancer. 2023;23(3):135–155. 10.1038/s41568-022-00541-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Broadinstitute/picard.[Java], Broad Institute. (Original work published 2014),2024. Reference Source [Google Scholar]
  10. Chen MJM, Chen LK, Lai YS, et al. : Integrating RNA-seq and ChIP-seq data to characterize long non-coding RNAs in Drosophila melanogaster. BMC Genomics. 2016;17(1): 220. 10.1186/s12864-016-2457-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chen Z, Duan X: Ribosomal RNA depletion for massively parallel bacterial RNA-sequencing applications.In: Y. M. Kwon & S. C. Ricke (Eds.), High-Throughput Next Generation Sequencing: Methods and Applications, Methods Mol Biol. Humana Press,2011;733:93–103. 10.1007/978-1-61779-089-8_7 [DOI] [PubMed] [Google Scholar]
  12. Cui P, Lin Q, Ding F, et al. : A comparison between ribo-minus RNA-sequencing and polyA-selected RNA-sequencing. Genomics. 2010;96(5):259–265. 10.1016/j.ygeno.2010.07.010 [DOI] [PubMed] [Google Scholar]
  13. Culviner PH, Guegler CK, Laub MT: A simple, cost-effective, and robust method for rRNA depletion in RNA-sequencing studies. mBio. 2020;11(2):e00010–20. 10.1128/mBio.00010-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Daines B, Wang H, Wang L, et al. : The Drosophila melanogaster transcriptome by paired-end RNA sequencing. Genome Res. 2011;21(2):315–324. 10.1101/gr.107854.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dawid IB, Wellauer PK: Ribosomal DNA and related sequences in Drosophila melanogaster. Cold Spring Harb Symp Quant Biol. 1978;42 Pt 2:1185–1194. 10.1101/sqb.1978.042.01.119 [DOI] [PubMed] [Google Scholar]
  16. de Queiroz BR, Laghrissi H, Rajeev S, et al. : Axonal RNA localization is essential for long-term memory. Nat Commun. 2025;16(1): 2560. 10.1038/s41467-025-57651-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Deng ZL, Münch PC, Mreches R, et al. : Rapid and accurate identification of ribosomal RNA sequences via deep learning. Nucleic Acids Res. 2022;50(10): e60. 10.1093/nar/gkac112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Dobin A, Davis CA, Schlesinger F, et al. : STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. 10.1093/bioinformatics/bts635 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fortini ME, Skupski MP, Boguski MS, et al. : A survey of human disease gene counterparts in the Drosophila genome. J Cell Biol. 2000;150(2):F23–F30. 10.1083/jcb.150.2.f23 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Fowler EK, Mohorianu I, Smith DT, et al. : Small RNA populations revealed by blocking rRNA fragments in Drosophila melanogaster reproductive tissues. PLoS One. 2018;13(2): e0191966. 10.1371/journal.pone.0191966 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Glover DM, Hogness DS: A novel arrangement of the 18S and 28S sequences in a repeating unit of Drosophila melanogaster rDNA. Cell. 1977;10(2):167–176. 10.1016/0092-8674(77)90212-4 [DOI] [PubMed] [Google Scholar]
  22. Goodman LD, Bellen HJ: Recent insights into the role of glia and oxidative stress in Alzheimer’s disease gained from Drosophila. Curr Opin Neurobiol. 2022;72:32–38. 10.1016/j.conb.2021.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hardin PE, Hall JC, Rosbash M: Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels. Nature. 1990;343(6258):536–540. 10.1038/343536a0 [DOI] [PubMed] [Google Scholar]
  24. Harvey KF, Pfleger CM, Hariharan IK: The Drosophila Mst ortholog, hippo, restricts growth and cell proliferation and promotes apoptosis. Cell. 2003;114(4):457–467. 10.1016/s0092-8674(03)00557-9 [DOI] [PubMed] [Google Scholar]
  25. Haugen RJ, Barnier C, Elrod ND, et al. : Regulation of the Drosophila transcriptome by Pumilio and the CCR4–NOT deadenylase complex. RNA. 2024;30(7):866–890. 10.1261/rna.079813.123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Huang R, Jaritz M, Guenzl P, et al. : An RNA-Seq strategy to detect the complete coding and non-coding transcriptome including full-length imprinted macro ncRNAs. PLoS One. 2011;6(11): e27288. 10.1371/journal.pone.0027288 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hughes ME, Grant GR, Paquin C, et al. : Deep sequencing the circadian and diurnal transcriptome of Drosophila brain. Genome Res. 2012;22(7):1266–1281. 10.1101/gr.128876.111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kaneko S, Miyoshi K, Tomuro K, et al. : Mettl1-dependent m 7G tRNA modification is essential for maintaining spermatogenesis and fertility in Drosophila melanogaster. Nat Commun. 2024;15(1): 8147. 10.1038/s41467-024-52389-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kellman BP, Baghdassarian HM, Pramparo T, et al. : Multiple freeze-thaw cycles lead to a loss of consistency in poly(A)-enriched RNA sequencing. BMC Genomics. 2021;22(1): 69. 10.1186/s12864-021-07381-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kidd SJ, Glover DM: Drosophila melanogaster ribosomal DNA containing type II insertions is variably transcribed in different strains and tissues. J Mol Biol. 1981;151(4):645–662. 10.1016/0022-2836(81)90428-9 [DOI] [PubMed] [Google Scholar]
  31. Kim SK, Tsao DD, Suh GSB, et al. : Discovering signaling mechanisms governing metabolism and metabolic diseases with Drosophila. Cell Metab. 2021;33(7):1279–1292. 10.1016/j.cmet.2021.05.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kissopoulou A, Jonasson J, Lindahl TL, et al. : Next generation sequencing analysis of human platelet PolyA+ mRNAs and rRNA-depleted total RNA. PLoS One. 2013;8(12): e81809. 10.1371/journal.pone.0081809 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kolde R: pheatmap: Pretty Heatmaps (p. 1.0.12).[Dataset].2010. 10.32614/CRAN.package.pheatmap [DOI]
  34. Koppaka O, Bakthavachalu B, Tandon S, et al. : Efficient ribosomal RNA depletion from Drosophila total RNA for next-generation sequencing applications.2024b. 10.6084/m9.figshare.28093562.v1 [DOI] [Google Scholar]
  35. Koppaka O, Tandon S, Chodankar A, et al. : Efficient ribosomal RNA depletion from Drosophila total RNA for next-generation sequencing applications. bioRxiv. 2024a; 2024.11.28.625868. 10.1101/2024.11.28.625868 [DOI] [Google Scholar]
  36. Kraus AJ, Brink BG, Siegel TN: Efficient and specific oligo-based depletion of rRNA. Sci Rep. 2019;9(1): 12281. 10.1038/s41598-019-48692-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lefebvre FA, Benoit Bouvrette LP, Bergalet J, et al. : Biochemical fractionation of time-resolved Drosophila embryos reveals similar transcriptomic alterations in replication checkpoint and histone mRNA processing mutants. J Mol Biol. 2017;429(21):3264–3279. 10.1016/j.jmb.2017.01.022 [DOI] [PubMed] [Google Scholar]
  38. Liao Y, Smyth GK, Shi W: featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–930. 10.1093/bioinformatics/btt656 [DOI] [PubMed] [Google Scholar]
  39. Lim SJ, Boyle PJ, Chinen M, et al. : Genome-wide localization of exosome components to active promoters and chromatin insulators in Drosophila. Nucleic Acids Res. 2013;41(5):2963–2980. 10.1093/nar/gkt037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Link N, Bellen HJ: Using Drosophila to drive the diagnosis and understand the mechanisms of rare human diseases. Development. 2020;147(21): dev191411. 10.1242/dev.191411 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Love MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12): 550. 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Malwade A, Koner A, Vibishan B, et al. : Experimental evolution of Drosophila without its microbiome. bioRxiv. 2024; 2024.10.25.620247. 10.1101/2024.10.25.620247 [DOI] [Google Scholar]
  43. Mirzoyan Z, Sollazzo M, Allocca M, et al. : Drosophila melanogaster: a model organism to study cancer. Front Genet. 2019;10:51. 10.3389/fgene.2019.00051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Mugat B, Akkouche A, Serrano V, et al. : MicroRNA-dependent transcriptional silencing of transposable elements in Drosophila follicle cells. PLoS Genet. 2015;11(5): e1005194. 10.1371/journal.pgen.1005194 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Nemeth K, Bayraktar R, Ferracin M, et al. : Non-coding RNAs in disease: from mechanisms to therapeutics. Nat Rev Genet. 2024;25(3):211–232. 10.1038/s41576-023-00662-1 [DOI] [PubMed] [Google Scholar]
  46. Nüsslein-Volhard C, Wieschaus E: Mutations affecting segment number and polarity in Drosophila. Nature. 1980;287(5785):795–801. 10.1038/287795a0 [DOI] [PubMed] [Google Scholar]
  47. O’Neil D, Glowatz H, Schlumpberger M: Ribosomal RNA depletion for efficient use of RNA-seq capacity. Curr Protoc Mol Biol. 2013;Chapter 4:Unit 4.19. 10.1002/0471142727.mb0419s103 [DOI] [PubMed] [Google Scholar]
  48. Pellegrini M, Manning J, Davidson N: Sequence arrangement of the rDNA of Drosophila melanogaster. Cell. 1977;10(2):213–224. 10.1016/0092-8674(77)90215-x [DOI] [PubMed] [Google Scholar]
  49. Pritykin Y, Brito T, Schupbach T, et al. : Integrative analysis unveils new functions for the Drosophila Cutoff protein in noncoding RNA biogenesis and gene regulation. RNA. 2017;23(7):1097–1109. 10.1261/rna.058594.116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Singer M, Berg P: Genes & genomes: a changing perspective.University Science Books,1991. Reference Source [Google Scholar]
  51. Souidi A, Jagla K: Drosophila heart as a model for cardiac development and diseases. Cells. 2021;10(11):3078. 10.3390/cells10113078 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sweeney BA, Petrov AI, Burkov B, et al. : RNAcentral: a hub of information for non-coding RNA sequences. Nucleic Acids Res. 2019;47(D1):D221–D229. 10.1093/nar/gky1034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Talross GJS, Carlson JR: The rich non-coding RNA landscape of the Drosophila antenna. Cell Rep. 2023;42(5): 112482. 10.1016/j.celrep.2023.112482 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Tang F, Barbacioru C, Wang Y, et al. : mRNA-Seq whole-transcriptome analysis of a single cell. Nat Methods. 2009;6(5):377–382. 10.1038/nmeth.1315 [DOI] [PubMed] [Google Scholar]
  55. Tariq MA, Kim HJ, Jejelowo O, et al. : Whole-transcriptome RNAseq analysis from minute amount of total RNA. Nucleic Acids Res. 2011;39(18): e120. 10.1093/nar/gkr547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Tautz D, Hancock JM, Webb DA, et al. : Complete sequences of the rRNA genes of Drosophila melanogaster. Mol Biol Evol. 1988;5(4):366–376. 10.1093/oxfordjournals.molbev.a040500 [DOI] [PubMed] [Google Scholar]
  57. Thompson MK, Kiourlappou M, Davis I: Ribo-Pop: simple, cost-effective, and widely applicable ribosomal RNA depletion. RNA. 2020;26(11):1731–1742. 10.1261/rna.076562.120 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tolwinski NS: Introduction: Drosophila—a model system for developmental biology. J Dev Biol. 2017;5(3):9. 10.3390/jdb5030009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. van Lopik J, Alizada A, Trapotsi MA, et al. : Unistrand piRNA clusters are an evolutionarily conserved mechanism to suppress endogenous retroviruses across the Drosophila genus. Nat Commun. 2023;14(1): 7337. 10.1038/s41467-023-42787-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wahl A, Huptas C, Neuhaus K: Comparison of rRNA depletion methods for efficient bacterial mRNA sequencing. Sci Rep. 2022;12(1): 5765. 10.1038/s41598-022-09710-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Wang Z, Gerstein M, Snyder M: RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009;10(1):57–63. 10.1038/nrg2484 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. White RL, Hogness DS: R loop mapping of the 18S and 28S sequences in the long and short repeating units of Drosophila melanogaster rDNA. Cell. 1977;10(2):177–192. 10.1016/0092-8674(77)90213-6 [DOI] [PubMed] [Google Scholar]
  63. Winnebeck EC, Millar CD, Warman GR: Why does insect RNA look degraded? J Insect Sci. 2010;10(1): 159. 10.1673/031.010.14119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Yang L, Duff MO, Graveley BR, et al. : Genomewide characterization of non-polyadenylated RNAs. Genome Biol. 2011;12(2): R16. 10.1186/gb-2011-12-2-r16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Ye C, Ho R, Moberg KH, et al. : Adverse impact of female reproductive signaling on age-dependent neurodegeneration after mild head trauma in Drosophila. eLife. 2024;13: RP97908. 10.7554/eLife.97908 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zhao W, He X, Hoadley KA, et al. : Comparison of RNA-Seq by poly (A) capture, ribosomal RNA depletion, and DNA microarray for expression profiling. BMC Genomics. 2014;15(1): 419. 10.1186/1471-2164-15-419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zhao S, Zhang Y, Gamini R, et al. : Evaluation of two main RNA-seq approaches for gene quantification in clinical RNA sequencing: polyA+ selection versus rRNA depletion. Sci Rep. 2018;8(1): 4781. 10.1038/s41598-018-23226-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
Wellcome Open Res. 2025 May 29. doi: 10.21956/wellcomeopenres.26694.r123472

Reviewer response for version 2

Jean-Yves Roignant 1

I went through author's responses and I am satisfied with the revisions. I therefore approve the article without reservation.

Is the rationale for developing the new method (or application) clearly explained?

Yes

Is the description of the method technically sound?

Yes

Are the conclusions about the method and its performance adequately supported by the findings presented in the article?

Partly

If any results are presented, are all the source data underlying the results available to ensure full reproducibility?

Yes

Are sufficient details provided to allow replication of the method development and its use by others?

Yes

Reviewer Expertise:

NA

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Wellcome Open Res. 2025 May 27. doi: 10.21956/wellcomeopenres.26694.r123471

Reviewer response for version 2

Amaresh Chandra Panda 1

The manuscript has improved significantly after the revision. The authors have answered/revised the manuscript according to the suggestions. I have no further comments to make.

Is the rationale for developing the new method (or application) clearly explained?

Yes

Is the description of the method technically sound?

Partly

Are the conclusions about the method and its performance adequately supported by the findings presented in the article?

Partly

If any results are presented, are all the source data underlying the results available to ensure full reproducibility?

Yes

Are sufficient details provided to allow replication of the method development and its use by others?

Yes

Reviewer Expertise:

RNA sequencing, rRNA depletion, non-coding RNA

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Wellcome Open Res. 2025 Apr 16. doi: 10.21956/wellcomeopenres.25929.r121320

Reviewer response for version 1

Jean-Yves Roignant 1

In this manuscript the authors develop a method to deplete ribosomal RNA in Drosophila melanogaster. They claim that current kits are not optimized for Drosophila, and therefore new methods should be implemented.

I have not doubt that this would be a great addition for the drosophila community. Nevertheless the authors omitted to mention the kits from siTOOLs that were specifically developed for Drosophila, one for full RNA and another for degraded RNA. They should definitively mention these kits in TABLE1 and ideally compare these kits with their method.

The authors used Ribodetector, a new tool developed for removing rRNA reads which was trained in the human genome. The authors could check the rRNA reads using bowtie with the Drosophila genome to check if they would get similar results.

The authors mentioned the difference in mapped reads between the probe system and the kit system, it would be interesting to know the type of unmapped reads. Is it related to the difference of number of rRNA reads? It would be interesting to know if the number of mapped reads would change if the step of removing rRNA reads was performed before or after the mapping with STAR. Since normally, rRNA reads are removed before mapping but the authors perform it after mapping with STAR.

Is the rationale for developing the new method (or application) clearly explained?

Yes

Is the description of the method technically sound?

Yes

Are the conclusions about the method and its performance adequately supported by the findings presented in the article?

Partly

If any results are presented, are all the source data underlying the results available to ensure full reproducibility?

Yes

Are sufficient details provided to allow replication of the method development and its use by others?

Yes

Reviewer Expertise:

Drosophila, molecular biology

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Wellcome Open Res. 2025 May 1.
Baskar Bakthavachalu 1

We thank the reviewers for their thorough evaluation of our manuscript and the constructive feedback. The suggestions have been instrumental in improving the clarity of our data presentation and strengthening the rigor of the methods section. Below, we provide detailed responses to each of the reviewers’ comments.

1. Nevertheless, the authors omitted to mention the kits from siTOOLs that were specifically developed for Drosophila, one for full RNA and another for degraded RNA. They should definitively mention these kits in TABLE1 and ideally compare these kits with their method. We thank the reviewer for pointing this out and have now included the siTOOLs in the revised manuscript and in Table-1. However, a direct comparison will not be possible, as our study is focused specifically on depleting rRNA using custom probes, rather than conducting a comparative analysis of different methods.

2. The authors used Ribodetector, a new tool developed for removing rRNA reads which was trained in the human genome. The authors could check the rRNA reads using bowtie with the Drosophila genome to check if they would get similar results. While we acknowledge the reviewer's concern regarding the use of Ribodetector for estimating Drosophila rRNA reads, it is important to note that Ribodetector relies on the human genome solely for the estimation of false positives and false negatives. The tool was trained using sequences from the SILVA database, which does not specifically focus on human rRNA sequences. In fact, Ribodetector has been widely used for rRNA estimation in non-human genomes as well (Dyksma & Pester, 2023; Nenciarini et al., 2024; Zhang et al., 2024). To further address the reviewer's query, we constructed additional analysis using 1000 randomly selected 100 bp reads from the fly rRNA sequences and another 1000 sequences from the whole dm6 reference genome. Our analysis showed that Ribodetector was able to correctly identify 98.6% of rRNA sequences while detecting only one rRNA sequence out of a thousand from the dm6 genome suggesting the tool is compatible for use with Drosophila rRNA.

3. The authors mentioned the difference in mapped reads between the probe system and the kit system, it would be interesting to know the type of unmapped reads. Is it related to the difference of number of rRNA reads? It would be interesting to know if the number of mapped reads would change if the step of removing rRNA reads was performed before or after the mapping with STAR. Since normally, rRNA reads are removed before mapping but the authors perform it after mapping with STAR. Unmapped reads typically originate from various sources. During our quality assessment of sequencing data generated from probe-based rRNA depletion, using FASTQC, we observed duplication rates ranging between 7–12%. These duplicates are primarily introduced during library preparation and PCR amplification and are subsequently removed post-alignment using Picard Tools. Notably, multimapping rates exceeding 15% often indicate inadequate rRNA depletion; our samples remained below this threshold. Therefore, a multimapping rate of 7–12% is considered acceptable for libraries prepared using custom rRNA depletion probes (Dobin & Gingeras, 2015; Fu et al., 2018; Picard Tools - Broad Institute). Approximately 5–6% of the reads were classified as "too short," which could result from sample contamination or residual rRNA. However, given that poly(A)-selected libraries also show a similar proportion (~5%) of short reads, this fraction in our rRNA-depleted libraries is not considered significant. Additionally, we observed a small fraction (~0.4%) of reads that failed to align due to factors other than mismatches or short length, an expected and acceptable rate (Dobin & Gingeras, 2015). As recommended by the reviewer, we aligned reads from the custom probe-based rRNA-depleted libraries to rRNA sequences, extracted the unmapped reads, and subsequently aligned them to the Drosophila melanogaster dm6 reference genome. This process resulted in only a marginal improvement in mapping efficiency, which is consistent with the low levels of rRNA contamination detected by Ribodetector. We applied the same approach to sequencing data from the kit-based rRNA depletion method where we notice ~7% of the reads maps to multiple loci, while ~57% of the reads are too short for mapping, suggesting potential issues at the step of rRNA depletion or cDNA synthesis. While we agree that a deeper characterization of the unmapped reads could be informative, we consider this beyond the scope of the current study. Nonetheless, those interested in further analysis can do so using the publicly available raw data under accession number GSE282990.

Wellcome Open Res. 2025 Mar 27. doi: 10.21956/wellcomeopenres.25929.r119970

Reviewer response for version 1

Amaresh Chandra Panda 1

In this manuscript, Koppaka & colleagues developed a cost-effective rRNA-depletion method for Drosophila. Here, they designed DNA probes against the rRNA and optimized the efficient digestion of the rRNA using RNase H. The authors claim that this method lead to better depletion of rRNA and enrichment of noncoding RNA in the drosophila RNA-seq library.

Although this is a fascinating study, I have a few comments to further improve the work.

  1. The number of RNA-seq reads in commercial kit samples is ~20% of the reads obtained in the custom probe sample. The higher percentage of rRNA reads and the low detection of non-coding RNAs in the commercial kit could be attributed to the sequencing depth. Similarly, in the polyA library, the read numbers are also very low compared to custom probe library. It might be possible that the buffer used for rRNA depletion by the commercial kit may be incompatible/inefficient for the cDNA library preparation by the library kit. Also, if there are any differences in the amount of libraries made from equal amounts of starting RNA, they may be discussed. 

  2. The major conclusions of the article are based on a single library prepared with custom probes compared to a library prepared from a commercial kit. In addition, there is a huge difference in the depth of sequencing as shown by the million reads.  It would be better to use at least two biological replicates with similar sequencing depth in RNA-seq to increase confidence in the differential expression analysis, as well as the differences observed in rRNA depletion efficiency.

  3. It might be better to mention the % of rRNA sequence covered by the probes. It might be better to discuss the fate of the portions of rRNA that are not targeted by the probes after digestion.

  4. The commercial kit used for comparison uses rRNA depletion by pulldown while the method used here used RNase H mediated digestion. Using a kit that use RNase H for rRNA depletion might have been a better comparison.

  5.  Table 2 mentions one poly (A) library with a custom probe. It is unclear whether the polyA selection was performed after rRNA depletion using a custom probe or a commercial kit for rRNA depletion. Please revise the methods section for a better understanding of samples and their comparisons. It might be better to include the sample table in the main text/Figure.

  6. The tapestation data for before and after rRNA depletion using a commercial kit may be provided.

Is the rationale for developing the new method (or application) clearly explained?

Yes

Is the description of the method technically sound?

Partly

Are the conclusions about the method and its performance adequately supported by the findings presented in the article?

Partly

If any results are presented, are all the source data underlying the results available to ensure full reproducibility?

Yes

Are sufficient details provided to allow replication of the method development and its use by others?

Yes

Reviewer Expertise:

RNA sequencing, rRNA depletion, non-coding RNA

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Wellcome Open Res. 2025 May 1.
Baskar Bakthavachalu 1

We thank the reviewer for his careful evaluation of our manuscript and for the constructive feedback provided. The comments have been valuable in enhancing the clarity of our data presentation and the rigor of the methods section. Below, we address each of the reviewers’ queries in detail.

1. The number of RNA-seq reads in commercial kit samples is ~20% of the reads obtained in the custom probe sample. The higher percentage of rRNA reads and the low detection of non-coding RNAs in the commercial kit could be attributed to the sequencing depth. Similarly, in the polyA library, the read numbers are also very low compared to custom probe library. It might be possible that the buffer used for rRNA depletion by the commercial kit may be incompatible/inefficient for the cDNA library preparation by the library kit. Also, if there are any differences in the number of libraries made from equal amounts of starting RNA, they may be discussed. 

Response: This study was designed to demonstrate an effective rRNA depletion method. The method was compared with a kit that was available with us. Normally, a sequencing depth of around 10 million is sufficient to perform differential expression analysis (Liu et al., 2014). The Drosophila melanogaster transcriptome is roughly 60Mb and a total bases needed for 20X coverage will be about 1.2 Gb. So, about 6 million paired-end reads of 200 bp will be sufficient to identify most transcripts. Both our custom probe-based method as well as kit-based rRNA depleted samples have reads beyond this estimated depth. We obtained 25 million paired-end reads for kit based rRNA depletion, while about 11 million paired-end reads for the same sample that was enriched using PolyA beads which is well within the acceptable range. In addition, in our original submission, we seemed to have accidentally mentioned just the single end read depth data, but in actual, we had used the double that read depth which is now corrected in the supplementary data Table 2 of the revised submission. It is also available under accession number GSE282990. QIAseq® FastSelect™ Handbook has a section that describes cDNA synthesis and library preparation using NEBNext Ultra II Directional Library Prep Kit and so it is unlikely that the reagents of the rRNA depletion kit is incompatible with cDNA library preparation. That said, though we obtained a relatively lower number of reads for PolyA-enriched samples, we found better mapping percentages as compared to rRNA depletion, suggesting this could not be the reason for the poor performance of the kit.  All libraries were prepared using 100 ng of total RNA for poly(A)-selection method and 1 µg of total RNA for rRNA depletion. The custom probe method produced reproducible results across both replicates, and this has now been included in the methods and discussion section.

2. The major conclusions of the article are based on a single library prepared with custom probes compared to a library prepared from a commercial kit. In addition, there is a huge difference in the depth of sequencing as shown by the million reads.  It would be better to use at least two biological replicates with similar sequencing depth in RNA-seq to increase confidence in the differential expression analysis, as well as the differences observed in rRNA depletion efficiency.

Response: We agree with the reviewer that at least two biological replicates are essential for the custom probe to be an acceptable method of choice for rRNA depletion. We had actually performed the experiments with two biological replicates. The data from both replicates are comparable, as shown in the revised supplementary table 2 and available under accession number GSE282990. The confusion regarding the use of a single library may have arisen because we initially reported the average sequencing depth across samples. We have now corrected this, and the updated table reflects the sequencing depth of each replicate individually. Furthermore, differential expression analysis was conducted between the two replicates of the PolyA-enriched samples and the two replicates of the rRNA-depleted samples, each with equivalent sequencing depth.

3. It might be better to mention the % of rRNA sequence covered by the probes. It might be better to discuss the fate of the portions of rRNA that are not targeted by the probes after digestion.

Response: Approximately 98% of the rRNA sequences (5S, 5.8S, 18S, and 28S) were covered by our probes. This has been explicitly mentioned in the revised manuscript. The regions not covered by the probes were extremely short. They would be eliminated during bead-based RNA cleanup following the RNAse H treatment.

4. The commercial kit used for comparison uses rRNA depletion by pulldown while the method used here used RNase H mediated digestion. Using a kit that uses RNase H for rRNA depletion might have been a better comparison.

Response: The Qiagen FastSelect Fly rRNA depletion kit does not employ a pulldown-based method. Instead, it uses oligos to inhibit reverse transcription into cDNA, thereby preventing rRNA sequences from being incorporated into the final sequencing library. To our knowledge, there is currently no commercially available rRNA depletion kit that utilizes an RNase H-based method for rRNA depletion in Drosophila. Finally, we would like to emphasize that the primary objective of this study is to demonstrate the efficiency of the custom probes in depleting rRNA, rather than to conduct a comparative analysis against existing commercial kits.

5. Table 2 mentions one poly (A) library with a custom probe. It is unclear whether the polyA selection was performed after rRNA depletion using a custom probe or a commercial kit for rRNA depletion. Please revise the methods section for a better understanding of samples and their comparisons. It might be better to include the sample table in the main text/Figure.

Response: In our method, total RNA was extracted and subsequently divided for PolyA enrichment and rRNA depletion. Two biological replicates were prepared for each condition: PolyA-enriched samples and rRNA-depleted samples using custom probes. We have now revised Table 2 and the methods section to more accurately reflect these steps.

6. The tapestation data for before and after rRNA depletion using a commercial kit may be provided.

Response: In RNA sequencing workflows, RNA is typically converted into cDNA, which is then used for sequencing. The Qiagen FastSelect Fly rRNA kit specifically inhibits reverse transcription of rRNA. Consequently, while rRNA remains physically present in the reaction mixture, it does not advance to later stages of library preparation. As a result, rRNA can still be detected by quality control systems such as the Tapestation, even when depletion has been effective resulting in visualization of peaks corresponding to rRNA bands. Thus, the efficacy of rRNA depletion cannot be accurately assessed based on Tapestation results. This is also explicitly mentioned in page 12 of QIAseq® FastSelect™ Handbook - “It is not possible to test the efficiency of the FastSelect reaction by running a portion of the eluate from the bead cleanup on a Bioanalyzer®, TapeStation®, Fragment Analyzer™, etc. FastSelect works by inhibiting reverse transcription of rRNA, which does not occur until the first-strand synthesis reaction during library prep.”

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Data Citations

    1. Kolde R: pheatmap: Pretty Heatmaps (p. 1.0.12).[Dataset].2010. 10.32614/CRAN.package.pheatmap [DOI]

    Data Availability Statement

    Underlying data

    NCBI Gene Expression Omnibus: RNA Sequencing data from NCBI GEO. Accession ID: GSE282990; http://identifiers.org/geo:GSE282990 ( Koppaka et al., 2024a).

    Extended data

    The extended data from this study can be accessed from:

    Figshare: Efficient ribosomal RNA depletion from Drosophila total RNA for next generation sequencing applications. https://doi.org/10.6084/m9.figshare.28093562 ( Koppaka et al., 2024b).

    The project contains the following extended data:

    • -

      Table S1: List of custom probes used for RNase H mediated rRNA-depletion

    • -

      Table S2: NGS data mapping details

    • -

      Table S3: Differential Expression analysis results

    • -

      Table S4: List of software used in the study

    Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

    Articles from Wellcome Open Research are provided here courtesy of The Wellcome Trust

    RESOURCES