Table 3 |.
Step | Problem | Possible Reason | Solution |
---|---|---|---|
37 | Insufficient colonies after library transformation (less than 300x library complexity) | Inefficient golden gate reactions and/or electroporation | Change to a fresh batch of reagents (T7 ligase, Esp3I, and electrocompetent cells are most critical) |
Confirm transformation efficiency with pUC19 control plasmid | |||
Input DNA mass for electroporation too low. | Increase the number of golden gate reactions to increase input DNA mass (see Step 5), carefully purify assembled DNA and quantify with Qubit prior to transformation | ||
45 | Representation of sgRNAs in plasmid library is skewed (skew ratio > 10), many expected sgRNA sequences are missing or are present in the wrong subpool | Jackpot effect from PCR amplification of oligo pool | Run multiple oligo pool PCR amplifications, use qPCR to ensure the reactions are halted during the exponential amplification phase, and pool the PCR products |
Inefficient golden gate reactions and/or electroporation | Follow troubleshooting solutions for Step 37 | ||
Insufficient representation (<300x) of library elements as input to any earlier step (oligo pool PCR, golden gate reactions, electroporation, NGS library prep) | Increase the number of oligo pool PCR and golden gate reactions (see Step 5), ensure adequate DNA mass input for electroporation, and increase the coverage of NGS reads per library element (see Step 44) | ||
Oligo pool library design errors | Ensure the recommended subpool dialout primers are used and these match the designed oligo PCR handle sequences | ||
71 | Titer of lentivirus batch is low | Transfection of and/or virus production by HEK 293FT cells inefficient | Use a fresh batch of low-passage HEK 293FT cells and replace transfection reagents |
Verify plasmid transfection and viral transduction efficiencies with a high-titer control fluorescent protein vector (e.g., Addgene #25999) by fluorescence imaging or flow cytometry | |||
76 | Yield of successfully transduced cells is low compared to the library complexity | Inefficient viral transduction | Make a fresh batch of lentivirus with improved titer or increase the spinfection scale to achieve the necessary library representation (see Steps 72–73) |
Miscalculation of viral titer | Re-titer the lentivirus batch and make a new batch if titer is low | ||
84 | Representation of sgRNA sequences in the cell library is skewed (skew ratio > 10) | Input plasmid library was skewed | Compare cell library skew to plasmid library skew and address as described in troubleshooting for Step 45 |
Spinfection inefficient or of insufficient scale | Follow troubleshooting solutions for Step 76 and ensure target representation of >300 cells/perturbation in spinfection scale calculation (see Steps 72–73) | ||
sgRNA sequences targeting essentials genes have decreased representation as compared to the plasmid library | Cells carrying sgRNAs targeting essential genes are dying | Filter out depleted guides or genes in screen analysis. If essential genes are important for a specific application, use an inducible Cas9 system to allow generation of the cell library prior to gene perturbation and optimize the time between Cas9 induction and screen readout. If essential gene depletion occurs prior to induction, inducible expression may be leaky and the cell line should be re-validated with control sgRNAs | |
86 | Plating of cells on imaging plates is uneven or too dense | Cells are not fully resuspended after trypsinization | Minimize cell clumps by optimizing enzymatic detachment, pipette-mixing the cell suspension, and filtering through a cell strainer if necessary |
Too many cells are seeded per well | Test multiple cell plating densities and methods of plating (e.g., gently swirling plate in a “figure 8” motion during plating). The best density and method for plating may vary by cell type, but poor plating can result in many problems such as poor phenotype images and/or in situ sequencing quality. | ||
Cells are not distributed evenly across the wells during plating | |||
95, 101, 103, 108, 110, 124, 126 | Evaporation during heated enzymatic or washing steps | Plate not fully sealed | Carefully seal the plate with a fresh foil seal before each incubation |
Fill empty wells and spaces between wells with water to limit evaporation | |||
Place a flat and moderately heavy item on top of the plate to keep the seal from detaching | |||
114 | Low sequencing spot brightness in the first cycle of SBS | RCA reaction inefficient or too short | Use a fresh batch of Phi29 polymerase and buffer; Phi29 polymerase is thermo-sensitive and should be stored appropriately, freeze-thaw cycles minimized, and RCA reaction mix assembled quickly on ice |
Run the RCA reaction for at least 16 hours | |||
Cell plating is too dense | Test at lower cell densities and see troubleshooting for step 86. | ||
High SBS background in the first cycle of SBS | Non-specifically bound dye-labeled nucleotides | Repeat post-incorporation washes (Step 110), proceeding rapidly between steps as demonstrated in the protocol video (https://youtu.be/TEqMbMjS1tA) | |
Background from phenotype stains | Image sequencing channels prior to first cycle incorporation to confirm origin of background and consider altering phenotyping approach following guidelines in Box 3 to reduce background | ||
High SBS dye background within cells in any cycle of sequencing (especially the “C” channel) | Insufficient washing before and/or after incorporation steps | Repeat heated post-incorporation washes (Step 110) and increase the number and duration of these washes for future cycles and plates | |
Wells drying out during solution exchanges | Carefully avoid plate drying by performing buffer exchange quickly | ||
Irreversible sticking of dye to cells or build-up of background over many cycles | Eliminate imaging of the DAPI channel in cycles after cycle 1 to avoid UV exposure-associated damage to DNA, using the SBS stain background to align subsequent cycles. The extent of background staining may vary by cell-type | ||
Incorrect sample storage | Store plate at 4C between SBS cycles (stable for weeks) | ||
Many sequencing spot appear to be outside of the boundaries of any cell | Diffusion of barcode mRNA or DNA from improper fixation, post-fixation, or a cell type-specific effect | Use fresh batches of PFA and glutaraldehyde | |
Ensure an LNA-modified reverse transcription primer is used | |||
While fixation and post-fixation conditions were uniformly applicable to the tested adherent cell lines, different sample types may require alternative strategies to fix cDNA to cell matrix such as biotin-streptavidin linkage | |||
Subpopulation of cells without sequencing spots | Incomplete antibiotic selection after lentiviral infection with perturbation library | Optimize concentration and duration of antibiotic selection using fresh antibiotic and parental cell line | |
Repeat selection steps with fresh antibiotic or re-infect and select starting from the parental cell line | |||
Highly variable sequencing spot counts within or between wells | Evaporation during earlier steps, especially after ethanol permeabilization | In addition to troubleshooting solutions for Step 95, ensure reagent exchanges in the well plate are completed quickly to avoid dehydration of samples | |
review proper ethanol removal procedure after permeabilization (Step 92) in the protocol video (see 06:38 at https://youtu.be/TEqMbMjS1tA). | |||
Regions of high cell density | Follow troubleshooting solutions for Step 86 to reduce cell density and/or uniformity | ||
Low average sequencing spot count per cell (<= 2) | Inefficient in situ enzymatic reactions | Use fresh batches of reagents (reverse transcription and gap-fill enzymes, dNTPs, PFA and glutaraldehyde) | |
Excess dNTPs during gap-fill reaction | Increase PBS-T washes prior to gap-fill reaction and ensure dNTPs are properly diluted | ||
Cell plating is too dense | See troubleshooting solutions for Step 86 and optimize cell density to avoid over-confluency | ||
Sequencing spot count may vary by cell type | If an un-validated cell line is being used, barcode expression may be lower or full protocol may need to be re-optimized, including choice of promoter or by using a CROPseq vector that incorporates a fluorescent protein and flow-sorting high-expressing cells (see Box 1) | ||
Decrease in spot fluorescence intensity over sequencing cycles | Photodamage from excess DAPI exposure | Use the minimum DAPI exposure (excitation intensity and duration) needed to acquire alignable images. If possible, avoid imaging DAPI at magnification higher than 10X after RCA. | |
Some fields-of-view out of focus | Adaptive focus control not tracking focal plane across the plate due to well plate quality, imaging solution volume, or because the bottom of the well plate is dirty | Increase volume of imaging media as some infrared autofocus systems may confuse the solution meniscus for plate surface | |
Thoroughly and carefully clean the bottom of the well plate using lens paper and isopropyl alcohol, allow to air dry | |||
Check for plate flatness and use recommended glass-bottomed plates if possible | |||
Contact microscope representatives for system-specific help | |||
Fields-of-view from successive cycles of sequencing are misaligned by >10% of the field-of-view size | Incorrect plate alignment during plate loading | Practice carefully aligning the plate to a standard location each time it is loaded onto the stage; use a spring-clamped plate holder to ensure minimal displacement | |
Microscope control software issues | Contact relevant microscope representatives for system-specific help | ||
Poor stage repeatability | Test the ability of the stage to return to the same location and contact relevant microscope representatives for system-specific help | ||
145, 150 | Rate of in situ sequencing reads mapping to expected barcode sequences is low (<70%) | Sequencing spot brightness too low | Inspect sequencing images and follow relevant troubleshooting solutions from Step 114 |
SBS dye background staining too high | |||
Fields-of-view from successive cycles of sequencing are not aligned | |||
Microscope setup not sufficiently separating SBS dye emission | Validate successful read mapping using a given microscope configuration on a technical experiment (e.g., frameshift reporter43, also see Box 2) | ||
150 | Many cells have reads mapping to multiple expected barcode sequences | MOI for sgRNA lentiviral infection is too high |
Make a fresh batch of lentivirus or re-titer the current batch |
Restrict screen analysis to cells mapping to only a single barcode (if MOI is very high and spot count and/or read mapping rate is low, this can lead to many false-positive cells) | |||
Poor cell segmentation | Adjust cell segmentation parameters, compare morphological and CellPose segmentation results, or use an algorithm optimized for a given cell type or experimental condition |