Genetic enhancement of climate-resilient traits in small millets: A review

Abstract

Agriculture is facing the challenge of feeding the ever-growing population that is projected to reach ten billion by 2050. While improving crop yield and productivity can address this challenge, the increasing effects of global warming and climate change seriously threaten agricultural productivity. Thus, genomics and genome modification technologies are crucial to improving climate-resilient traits to enable sustained yield and productivity; however, significant research focuses on staple crops such as rice, wheat, and maize. Crops that are naturally climate-resilient and nutritionally superior to staple cereals, such as small millets, remain neglected and underutilized by mainstream research. The ability of small millets to grow in marginal regions having limited irrigation and poor soil fertility makes these crops a better choice for cultivation in arid and semi-arid areas. Hence, mainstreaming small millets for cultivation and using omics technologies to dissect the climate-resilient traits to identify the molecular determinants underlying these traits are imperative for addressing food and nutritional security. In this context, the review discusses the genomics and genome modification approaches for dissecting key traits in small millets and their application for improving these traits in cultivated germplasm. The review also discusses biofortification for nutritional security and machine-learning approaches for trait improvement in small millets. Altogether, the review provides a roadmap for the effective use of next-generation approaches for trait improvement in small millets. This will lead to the development of improved varieties for addressing multiple insecurities prevailing in the present climate change scenario.

1. Introduction

Global food security and nutritional deficiencies are the two most challenging situations for the growing human population. Crop diversification helps to a great extent to achieve food security [1]. However, changing climatic conditions significantly influence food security by limiting crop production [2]. Globally, malnutrition or hidden hunger is another major problem increasing substantially in developing countries. The annual monocrops dominating the current agricultural system, such as wheat (Triticum spp.), rice (Oryza sativa), and maize (Zea mays), are calorie-dense and require extensive inputs of pesticides and fertilizers to maintain the yield every year [3]. Additionally, human populations inhabiting Africa and Asia lack access to different food crops, and the staple crops consumed by them fall short of supplying essential micronutrients, which is the prime cause of malnutrition [4]. In this context, crop diversification is the most effective approach to improve resistance against biotic factors and tolerance to abiotic stressors, leading to food and income security for farmers [5].

Diversifying mainstream staple crops with other crops would help achieve nutritional security and sustainable agriculture goals. Considering this, millets, also known as nutri-cereals, have the potential to provide adequate macro- and micro-nutrients compared to staple crops. Millets are a rich source of dietary fibers, essential amino acids and minerals, antioxidants, trace elements, protein, fats, and carbohydrates [6]. Additionally, millets have exceptional climate resilience traits, enabling them to survive and grow in various climatic and soil conditions [7]. Their excellent nutritional profile and climate resilience features make millets known as miracle grains [1]. Despite numerous superior qualities, millets are underutilized due to the substantial focus on major cereal crops worldwide. Thus, millets are referred to as “orphan cereal crops” by the scientific community [8].

Developing novel biotechnological approaches and breeding technologies provides a platform for exploiting millets' genetic diversity and accelerating trait improvement programs. These tools have the potential to mine the gene pools of crops to identify the agronomically important traits in millet. Presently the genome sequence of several millet crops is available to hasten the progress of crop improvement programs in these crops [[9], [10], [11]]. Breeding techniques complemented by new genomic and molecular approaches offer a platform to identify candidate genes or markers associated with stress tolerance. Several omics studies have been performed in small millets to understand stress-related mechanisms [[11], [12], [13], [14], [15]]. Numerous candidate genes having roles in yield enhancement, stress resilience, and other important agronomic features have been identified in small millets [16,17]. For instance, EcHNRT2, EcLNRT1, EcNADH-NR, EcGS, EcFd-GOGAT, and EcDof1 have been identified in finger millet to develop NUE (Nitrogen use efficiency) varieties [18]. EcDehydrin7 involved in drought stress was identified in finger millet, which could be used to develop drought and heat-tolerant crops [19]. SiARDP and SiLEA14, involved in salt and drought tolerance, were functionally characterized in foxtail millet [20,21]. Still, they are waiting for further comprehensive functional characterization using advanced biotechnological tools. Novel genome editing tools, such as zinc-finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs) and CRISPR/Cas, have been developed to revolutionize the crop improvement approaches [22]. However, the use of these technologies in small millets is yet to be standardized.

Studies on different small millets have provided information about their phenotypic diversity, climate resilience features, genome-wide SNP markers, and the heritability of different important traits and markers for genetic diversity. However, these marker data have not been exploited successfully for improved variety development, and low yield remains a challenge for small millet cultivation. Currently, the advancement in genomics and genome modification technologies and computer-based approaches enables the application of novel tools for the varietal improvement of small millets. In this context, the review discusses current trends of next-generation tools and their application for small millet improvement. The review also discusses the success stories, future aspects, and potential risks of these techniques in breeding programs for improvement in small millets.

2. Importance of small millets and scope for their improvement

‘Small millets,’ also known as minor millets, is a generic name for coarse cereals, mainly cultivated in rain-fed areas of arid and semi-arid zones [8]. It includes foxtail millet (Setaria italica), barnyard millet (Echinochloa crusgalli), finger millet (Eleusine coracana), proso millet (Panicum miliaceum), little millet (Panicum sumatrense), kodo millet (Paspalum scrobiculatum), teff (Eragrostis tef), Job's tears (Coix lacryma-jobi), fonio (Digitaria exilis), browntop millet (Urochloa ramosa), and guinea millet (Brachiaria deflexa) [8,23]. Small millets belong to the Poaceae family but are superior to other major cereals in terms of nutritional qualities, climate resilience, and other agronomical traits. In addition, millets have excellent water-use and nitrogen-use efficiencies that help them survive under water-deficit conditions in rain-fed regions [24]. Given this, millets have a minimum dependency on irrigation and nitrogen fertilizers for better yield [25]. Due to their nutritional superiority and climate resilience features, small millets can supplement staple cereal crops. They are rich in micro- and macro-nutrients, proteins, essential amino acids, dietary fiber, and resistant starch (Table 1). In addition, small millets also have rich contents of phytochemicals, viz., alkaloids, flavonoids, phenolics, and saponins [26]. For example, kodo millet has the maximum content of phenolic compounds (10.3%), and finger millet is rich in reducing sugars (391.3 mg/g each) [26]. Ferulic acid, predominantly identified as free phenolic fractions, was recorded highest in kodo millet (99.35 mg/100 g), followed by finger millet and foxtail millet with 57.04 mg/100 g and 54.65 mg/100 g of ferulic acid, respectively [27]. Similarly, proso millet is the richest source of protein (12,610 mg/100 g) among millets [27]. Most small millets are gluten-free and have a low glycemic index [28]. However, several antinutrient factors like tannins, polyphenols and phytic acid present in millets limit the bioavailability of essential nutrients, hence limiting the consumption of millets [29]. It has been estimated that millets have 0.2–0.3% polyphenols, 0.48% phytates and 0.61% tannins, among which the presence of phytic acid is a matter of concern, as it affects the bio-accessibility of several major nutrients [30].

Table 1.

Nutritional profile of small millet grains in comparison to major cereals [28,112].

CROP	Protein (g/100 g)	Fibre (g/100 g)	Carbohydrate (g/100 g)	Thiamine (mg/100 g)	Riboflavin (mg/100 g)	Major and trace mineral elements (in mg/100 g)
						K	Ca	P	Mg	Cr	Zn	Fe	Cu	Mn
Barnyard millet	6.2	9.8	65.5	0.30	0.09	195	21	340	82	0.14	2.6	9.2	1.3	1.33
Finger millet	7.3	3.6	72.0	0.38	0.21	407.00	398	320	137	0.03	2.3	3.9	0.5	5.49
Foxtail millet	12.3	8.0	60.9	0.42	0.19	299.28	38	422	81	0.07	2.9	5.3	1.6	0.85
Kodo millet	8.3	9.0	65.9	0.33	0.10	181.785	31	215	166	0.08	1.5	3.6	5.8	2.9
Little millet	7.7	7.6	67.0	0.41	0.28	192.21	12	251	133	0.24	3.5	13.9	1.6	1.03
Pearl millet	11.8	2.3	67.0	0.38	0.21	275.675	46	379	137	0.02	3.1	8	1.1	1.15
Proso millet	12.5	7.2	70.4	0.59	0.11	177.315	23	281	117	0.04	2.4	4	5.8	1.2
Rice	11.8	0.2	78.2	0.41	0.04	35	21	433	177	-	6	2	0.2	0.974
Wheat	6.8	1.2	71.2	0.41	0.10	107	34	357	137	0.02	2.6	3.6	0.4	4.07
Sorghum	10.4	2.0	70.7	0.38	0.15	672	13.49	380	165	-	2.51	8.23	0.3	1.605

The cultivation of small millets is limited to small farmers of low-income countries due to the presence of various antinutrients and the limited exploitation of genetic resources for trait improvement. However, small millet cultivation provides grains for forage, enhances agricultural biodiversity, checks soil erosion in arid regions, and ensures higher carbon sequestration due to the C₄ mode of photosynthesis [8]. In recent years, foxtail millet has emerged as a model crop for C₄ plants among small millets. It, therefore, enables the breeding programs to explore complex agronomic traits in all C₄ millet crops [31]. Interestingly, recent research by Yang et al. [32] has confirmed the development of mini foxtail millet with a mutation, xiaomi, which has a short life cycle and flowering time. This makes the xiaomi mutant an Arabidopsis-like C₄ model plant. Except for foxtail millet, proso millet, and finger millet, inadequate germplasm conservation has been done for other small millets, which is a potential reason for limited trait improvement research programs on these crops [8]. Therefore, significant efforts must be made in other small millets to conserve and explore their germplasm before they are lost. The application of novel approaches is imperative to examine the agronomic traits followed by their improvement in small millets [33]. Trait improvement in small millets, particularly focusing on reducing antinutrients and improving yield, could promote their cultivation and consumption as nutri-cereals, which will help combat hidden hunger [33].

Germplasm resources of small millets are conserved in genebanks of different regions, led by Asian countries with 64.4% accessions [34]. Globally, germplasm resources of small millets are majorly conserved in the National Bureau of Plant Genetic Resources (NBPGR, India), Institute of Crop Science, Chinese Academy of Agricultural Sciences (ICS-CAAS), National Institute of Agrobiological Sciences, Japan, and Ethiopian Institute of Biodiversity [34]. The availability of genome sequence data of small millets can facilitate the development of mapping populations with desired traits. The availability of draft genome sequences for most millets is competent for large-scale genotyping and gene mining. For example, ICRISAT, in collaboration with Cornell University, has genotyped different small millets using genotyping-by-sequencing (GBS) and identified diversity in small millet population structure [35]. However, ploidy level and the presence of repetitive DNA greatly influence the success of genome sequencing in small millets. The emerging next-generation sequencing approaches complemented with extensive data analysis platforms are expected to provide great success in future. Further, the All India Coordinated Research Project on Small Millets (AICRP-Small millets) is actively performing small millet improvements through collaborative interaction of State Agricultural Universities, ICAR institutes and other voluntary centres [34]. Thus far, this coordinated effort has successfully released 248 varieties of small millets in India [36] (http://www.aicrpsm.res.in/). Prior to the genome sequence of small millets, molecular markers such as RAPD, RFLP, AFLP and SSRs were used to map genes associated with important traits. However, advanced genome sequencing approaches have enhanced the direct identification of genes. To date, several gene families have been identified and functionally characterized, such as MYB, NAC, WRKY, CDPK, WD40, AP2/ERF, ATG and LIM genes (Fig. 1) [[37], [38], [39], [40], [41], [42], [43]]. Additionally, several other stress-related genes have been identified, most of which have been studied in foxtail millet and finger millet. For example, SiLEA14 was significantly up-regulated in foxtail millet under osmotic and salt stress conditions [21]. Notably, seven drought stress-related proteins, namely, NAC2, U2-snRNP (small nuclear RiboNucleoProtein particles), CDPK, synaptotagmin, Aquaporins, MPK17-1 and Scythe protein, have been identified in small millets [44]. Further, EcDehydrin7, Ec-apx1, EcbHLH57, EcbZIP60, Ec-apx1, EcGBF3 and EcbZIP17 have been identified in finger millet, and their role in conferring stress tolerance has been studied in various model plants such as tobacco and Arabidopsis (Fig. 1) [45].

Fig. 1.

Overview of genes functionally characterized in small millets and their physiological functions under various stress conditions.

3. Factors limiting the yield and productivity of small millets

Due to their diverse climate resilience features, small millets are mainly grown in arid and semi-arid regions. Of the total area under millets (76 million ha), 18–20 million ha is utilized worldwide for small millets cultivation [34]. Though biotic and abiotic stresses least influence small millets, a significant variation exists among different germplasms for stress tolerance in small millets [46]. Therefore, identifying and applying agronomically important traits for millet breeding programs is imperative. Several abiotic stresses, namely, drought, salinity, heat, lodging, etc., and biotic stresses, viz., fungal and bacterial diseases, significantly influence the small millet productivity (Fig. 2).

Fig. 2.

Environmental stresses affecting small millets. Different abiotic and biotic factors that limits the yield and productivity of small millets is shown. Also, the anti-nutrients in millets are highlighted.

3.1. Water deficit stress

Water scarcity is the most eminent problem in arid and semi-arid regions of many countries, which leads to severe drought stress in those areas. On a report by the United Nations, approximately 40% of the world's land is dry, which feeds almost 2 billion people [47]. Notably, the severity and frequency of drought stress have enhanced hitherto [48]. As the dominant crops of these areas, small millets production is adversely affected by drought stress. Ajithkumar et al. (2014) [49] showed a negative impact of drought stress on yield, membrane integrity, osmotic balance, water potential maintenance, and photosynthetic efficiency in pearl millet. Similarly, teff production is limited by drought conditions, which needs immediate action to improve mitigation and adaptation strategies [50]. Further, Matsuura and the research group [51] examined the impact of water deficit conditions on the vegetative and reproductive stages of proso millet, foxtail millet, and little millet. The foxtail millet showed an 80% reduction in yield after drought treatment at the pre-heading stage, whereas proso millet and little millet showed a 36 and 20% yield reduction [51]. Although post-heading drought stress leads to severely reduced productivity in proso and little millet only, minor impact was observed in foxtail millet. Therefore, the study suggested that foxtail millet has better drought tolerance mechanisms than proso and little millet. Further, drought stress in finger millet resulted in complete yield loss [52], and almost 77% of yield loss was reported in teff after confronting drought stress at the heading stage of growth [53]. Winkel et al. (1997) [54] investigated the adverse effect of drought conditions on different stages of pearl millet. They concluded that yield loss is severe when it experiences drought at the reproductive stage. The yield loss in millets varies yearly based on the drought stress severity; however, millets produce grains even under severe drought conditions, unlike other major cereals.

Numerous agronomical traits are responsible for the drought tolerance features of small millets. Traits responsible for nitrogen-use efficiency, photosynthetic efficiency, early flowering, and rapid growth are important factors that help them escape drought conditions [55]. Other traits associated with water-use efficiency helps millets avoid drought stress by maintaining the water potential in cells [56]. Similarly, superior antioxidant qualities and osmolyte accumulation confer drought tolerance in millets [55]. In addition, several morphological, physiological, and biochemical traits have been identified in millets that are crucial for their drought tolerance. For example, Balsamo et al. (2006) [57] reported that bundle sheath lignification enhances teff tensile strength. Further, enhanced accumulation of osmolytes was observed in little millet under drought conditions [49]. Root architecture is another important morphological characteristic to survive drought conditions. Therefore, broader root architecture in teff allows it to survive drought stress [58,59]. Considering this, millets are widely grown in rain-fed regions of various developing countries.

3.2. Salinity stress

Salinity is a major problem in the irrigated area of agricultural lands, severely affecting crop yield and nutritional status. Currently, almost 45 mha of total irrigated land is affected by salinity [60]. Millets are also categorized as glycophytes as they are not resistant to high salt concentrations [61]. Given this, Rasool et al. (2020) [62] and Shah et al. (2020) [63] investigated the effect of salinity on foxtail millet. Due to salt stress, they reported reduced biomass, photosynthetic efficiency, and water potential. Finger millet is salt tolerant, ensuring grain production under saline conditions [64]. Several finger millet genotypes have been investigated for their salt tolerance ability, which reported significant genotype-based variation in terms of tolerance in finger millet [65]. Salt stress adversely affects germination and seedling development in finger millet [66]. Similarly, four foxtail millet genotypes were investigated for their salt tolerance and showed variable morphological and anatomical changes among the sensitive and tolerant cultivars [67]. Further, a study by Rahim et al. (2020) [68] assessed the salt tolerance ability of pearl millet and showed a reduction in the number of tillers and growth under high salt concentrations.

3.3. Temperature stress

Millets are heat tolerant; however, several molecular changes are induced under high temperatures. For example, photosynthetic efficiency, transpiration, and respiration are major biological processes that are sensitive to temperature and significantly impact yield [69]. Therefore, the annual rise in global temperature greatly influences crop productivity, which threatens food security. Finger millet cultivars are widely grown in areas with a maximum temperature of 36 °C; however, a further rise in temperature adversely affects yield [70]. The study reported significant variation in the germination of different germplasms of finger millet under increased temperature, suggesting the need for improved genotype development by breeding programs [70]. Further, pearl millet can produce grains at temperature up to 42 °C, where other crops fail to grow. However, changing climatic conditions and further increase in temperature will reduce the yield of pearl millet by 17% by 2050 [71]. Similarly, Knox et al. [72] estimated approximately a 10% loss in the yield of pearl millet in Asian regions. Djanaguiraman et al. [73] quantified the impact of high temperature on pollen, pistil, and germplasm variability.

3.4. Waterlogging stress

Waterlogging is a major stress in areas with heavy precipitation that adversely impacts stomatal closure, photosynthesis, and root architecture by inhibiting gas diffusion [74]. Waterlogging causes almost 16% of yield loss in proso millet [74]. Similarly, a reduction in the yield of finger millet was reported under waterlogging, as it favours an anaerobic mode of metabolism that is not very economical in terms of ATP production [75]. Matsuura et al. (2016) [76] investigated the effect of waterlogging on proso millet, little millet, and foxtail millet and suggested that little millet showed better waterlogging tolerance than other millets.

3.5. Lodging effect

Lodging, bending of the stem, is a severe problem in many small millet crops, which causes severe yield loss. Several climatic conditions, such as rain, wind, irrigation, or combined effect, can lead to lodging in millets. Lodging stress majorly influences the stem and root of the plant, hence regarded as stem lodging and root lodging, respectively [77]. Numerous studies have shown the negative impact of lodging stress on teff and foxtail millet. Teff is majorly susceptible to root lodging [78]. Similarly, lodging stress causes yield loss in foxtail millet [79]. Further, Opole et al. [80] showed the role of excessive fertilizer in causing lodging stress in finger millet, which severely reduces the yield. However, pearl millet showed tolerance to the lodging effect [81].

4. Biotic factors affecting small millets

Under vulnerable conditions, various biotic factors cause heavy losses in small millet production. Several pathogens such as Rhizoctonia solani, Pyricularia grisea, Pyricularia setariae, Ustilago crus-galli, Sphacelotheca destruens, Uromyces linearis, Uromyces eragrostidis, etc. have been reported in small millets [46] (Table 2). Among various pathogens, fungal diseases are more devastative for small millets, as they have broad spectrum of host. For example, smut and rust are commonly known to infect all the small millets including teff [46]. However, other diseases such as blast, leaf blight, udbatta, sheath rot, sheath blight and brown spot specifically infect one or few small millets [46]. Severe yield losses have been reported in small millets due to these pathogens. For example, 30–40% grain loss was reported in foxtail millet due to blast [82]. Similarly, blast disease causes over 50% loss of finger millet [83]. Blast disease caused by P. grisea, infects several small millets, including foxtail millet, finger millet, proso millet, barnyard millet and little millet. The average yield loss due to blast infection may extend to 90% in affected areas [84]. Further, teff rust is responsible for 10–40% yield losses annually [85]. Downy mildew of finger millet causes severe losses, with more than 50% yield loss in certain years [46]. Ragi mottle streak virus infects finger millet in India, where yield loss ranged from 50% to 100% in some affected regions [46]. Leaf blight is another common fungal disease of small millets. The fungi for blight disease in different small millets are Alternaria tenuissima (kodo millet), Cochliobolus setariae (foxtail millet), Drechslera nodulosa (little millet, tef), Exserohilum monoceras (barnyard millet), and Bipolaris panici-miliacei (proso millet) [46]. Several different fungi, such as Uromyces eragrostidis, Uromyces setariae-italicae, Uromyces linearis and Puccinia substriata are known to cause rust in small millets such as foxtail millet, finger millet, kodo millet and little millet [46]. Similarly, grain smut is a common disease of barnyard millet, little millet, finger millet and foxtail millet, whereas head smut majorly infects kodo millet, proso millet and teff [46]. Among small millets, no significant information is available about the disease of teff and fonio.

Table 2.

Common pathogens reported to cause diseases in small millets [184].

Crop	Disease	Pathogen
Finger millet	Blast	Pyricularia grisea
	Rust	Puccinia substriata
	Smut	Melanopsichium eleusinis
	Downy mildew	Sclerophthora macrospora
	Seedling & leaf blight	Drechslera nodulosum
	Cercospora leaf spot	Cercospora eleusinis
	Banded blight	Rhizoctonia solani
	Wilt or foot rot	Sclerotium rolfsii
	Bacterial leaf spot	Xanthomonas eleusinae
	Ragi severe mosaic	Sugarcane mosaic virus
	Ragi mottle streak	Ragi mottle streak virus
	Ragi streak	Maize streak virus (Eleusine strain)
Foxtail millet	Blast	Pyricularia setariae
	Rust	Uromyces setariae-italicae
	Smut	Ustilago crameri
	Downy mildew	Sclerospora graminicola
	Udbatta	Ephelis sp.
	Bacterial leaf blight	Pseudomonas avenae
Kodo millet	Head smut	Sorosporium paspali
	Rust	Puccinia substriata
	Udbatta	Ephelis sp
Barnyard millet	Head smut	Ustilago crus-galli
	Kernel smut	Ustilago panici-frumentacei
	Bacterial leaf blight	Pseudomonas avenae
Proso millet	Head smut	Sphacelotheca destruens
	Bacterial leaf blight	Pseudomonas avenae
Little millet	Rust	Uromyces linearis
Teff millet	Rust	Uromyces eragrostidis
	Damping off	Helminthosporium poae

5. Antinutrient content in small millet grains

Small millets are a rich source of various micro- and macro-nutrients, therefore considered important cereals to combat hidden hunger. However, different antinutrients such as tannins, polyphenols, and phytic acid reduce millets' nutritional qualities as they chelate various crucial cations like K⁺, Mg²⁺, Zn^2+, and Ca²⁺ in the human body after consumption [86]. This activity reduces the bioavailability of these cations to several important enzymes and hampers the biological process mediated by these enzymes. Additionally, various enzyme inhibitors, such as amylase and protease inhibitors, cause indigestion of millet grains [87]. These anti-nutritional qualities of millets are also important reasons for their orphan status and ignorance by agricultural communities. Further, these anti-nutrients also reduce the bioavailability of several nutrients and minerals, which leads to harmful effects on the health of people whose nutritional requirements solely depend on these crops [29]. Kumar et al. [88] reported the presence of various anti-nutrients such as oxalates, phytic acid, non-starch polysaccharides-glucans, tannins, and protease inhibitors in finger millet. Earlier, Ramachandra et al. [89] investigated the content of tannins in finger millet and identified it to be in the range of 0.04–3.47%. However, varietal variation of tannin content was observed, therefore paving the path for germplasm improvement. Similarly, protease inhibitor activity and non-starch polysaccharides contents have been identified in various finger millet varieties [90,91]. Pradeep and Sreerama [92] investigated the effect of various treatments on the anti-nutrient composition of barnyard millet, proso millet, and foxtail millet. Currently, several processing techniques, such as decortication, germination, milling, roasting, puffing, cooking, and hydrothermal treatments, are widely utilized to reduce the anti-nutrients in millets [93]. Numerous studies have reported the success of these techniques in different millets [[94], [95], [96], [97]]. However, these techniques also cause a significant reduction in other nutrients and minerals [95,98]. Therefore, genetic approaches to enhance the bioavailability of nutrients and reduce anti-nutrients are imperative for increasing these crops' popularity and consumption.

6. Genetic and genomic resources

Crop diversity is a primary requirement for sustainable agriculture and nutritional security. Diverse germplasm provides significant variability for crop improvement. Globally, approximately 133,849 accessions of small millets are conserved in genebanks (Supplementary Fig. S1) [34]. ICRISAT conserves around 10,193 germplasm accessions of all small millets from 50 countries [99]. Around 29,000 accessions of proso millet have been conserved worldwide, of which 849 germplasm are conserved at the International Crops Research Institute for Semi-Arid Tropics (ICRISAT) [100]. Over 8000 germplasm accessions each of kodo and barnyard millet, and around 3000 germplasm of little millet have been successfully stored in genebank [101]. Small millet crops are majorly tetraploid, except in foxtail millet and barnyard millet, which have diploid and hexaploidy levels, respectively [34]. These large collections of conserved small millet genotypes are grouped according to various parameters, such as region of origin and promising traits, by the genebank institutions like ICRISAT [99]. Each group represents a particular genetic diversity called “core collection” [99]. These core collections provide sufficient genetic diversity for crop improvement. The genetic diversity of small millets has been investigated with the help of several genetic markers, such as genome-wide small nucleotide polymorphism (SNP) markers (Table 3). This is facilitated by novel sequencing approaches such as genotyping by sequencing (GBS), restriction site-associated DNA (RAD) sequencing and whole genome resequencing [102]. Yue et al. (2016) [103] identified about 400,000 SNP markers and 35,000 simple-sequence repeats (SSRs) from proso millets. Similarly, random-amplification-of-polymorphic DNA (RAPD) markers have been identified in 96 kodo millet accessions [104]. Recently, the genotyping-by-sequencing (GBS) approach was used to identify novel SNP markers in 288 finger millet accessions from Zimbabwe and Ethiopia [105]. Around 2412 high-quality SNP markers were identified in proso millet, followed by a genome-wide association study to understand the genetic diversity among different accessions of proso millet from Western Europe, Eastern Asia, Americas, Southern Asia, Western Asia and Africa [106]. Further, 2977 SNP markers were used for a marker-trait association study in finger millet to understand the genetic diversity associated with seed protein content, grain yield and days to maturity [107]. Additionally, advanced phenomics methods could be integrated with genomics data to enhance the study of a diverse genetic pool of small millets. Thus, huge opportunities are available for researchers to unveil the genetic diversity of small millets, which certainly encourages crop improvement programs in millets.

Table 3.

Molecular markers developed for genotyping applications in small millets.

Crop	Markers	Methods	Major finding	Reference
Finger millet	Genic and genomic SSRs	Association mapping	Four QTLs for finger blast and one QTL for neck blast resistance were identified	[126]
	Genomic SSRs	Association mapping	Seven QTLs were associated with various agronomic traits including leaf blast resistance.	[128]
	23,000 single-nucleotide polymorphisms (SNPs)	Genotyping-by-sequencing	Genotyping-by-sequencing of 113 diverse finger millet accessions	[185]
	10,327 SSRs and 23,285 non-homeologous SNPs	Roche 454 and Illumina HiSeq 2000	identified new polymorphic SSRs and SNPs	[186]
	Whole-genome sequencing of cultivar ML-365	Whole genome and transcriptome	1196 Mb genome size covering 82% of crop genome. 85,243 genes were predicted and one half of the genome was found to be repetitive	[11]
	Assembled the genome of cultivar PR202 via a novel pipeline	Whole-genome NGS	size of the assembled genome was about 1.2 Gb and predicted 62,348 genes	[187]
Foxtail millet	highly polymorphic simple sequence repeat markers	genome-wide microsatellite variant analysis	733 highly polymorphic SSR loci	[188]
	EST-derived-SSR (eSSR) markers	–	447 eSSR markers developed in foxtail millet	[189]
	RAPD and ISSR markers	–	135 RAPD and 77 ISSR	[190]
	rhizoplane microbial biomarkers	GWAS, MWAS and mGWAS	257 rhizoplane microbial biomarkers	[191]
	SNPs	GBS-ddRAD approach	10 K SNPs	[113]
	SSRs	Genome-wide identification	74 SSRs	[192]
Kodo millet	SNPs	genotyping-by-sequencing (GBS)	3461 SNPs	[35]
Proso millet	SNPs	genotyping-by-sequencing (GBS)	1882 SNPs	[35]
	AFLP	AFLP method	514 polymorphic AFLP markers	[193]
Little millet	SSRs	–	4443 genic-SSR	[194]
	RAPD markers	RAPD analysis	175 RAPD marker	[195]
Barnyard millet	SSRs	–	51 SSR markers	[196]

7. Dissecting climate-resilient and nutritional traits in small millets

Millets exhibit several important nutritional and climate-resilient traits, which have been identified so far by using novel omics and breeding approaches. Modern genetic and genomic tools facilitate the identification of these traits. The availability of whole genome sequence information of several small millets has further accelerated the crop improvement programs in millets [8]. As climate-resilient crops, millets are underexplored in this regard, though the looming climate changes necessitate trait improvement practices to be deployed to avoid crop failure [108]. Both genomic and genetic resources are necessary for trait improvement, where genomic information accelerate the identification and functional characterization of trait-responsive genes for their utilization in breeding programs (Fig. 3; Supplementary Fig. S2). Among small millets, the whole genome sequence is available for foxtail millet, finger millet, barnyard millet, teff, and green foxtail [9,11,[109], [110], [111]] (Fig. 4; Supplementary Table S1). This breakthrough research expedites the identification of novel genes responsible for various nutritional and climate resilience traits in small millets (Table 4). The availability of genomic resources has accelerated breeding programs targeting trait improvement in small millets. Genome-wide association study identified ten loci associated with nutritional elements, including calcium, zinc, phosphorus, sulfur, magnesium, boron, nickel, potassium, and manganese [112]. GWAS, through the GBS-ddRAD approach, identified markers associated with various traits such as grain yield (GY), flag leaf width (FLW), thousand-grain weight (TGW), etc. [113]. The Pearl Millet Inbred Germplasm Association Panel (PMiGAP) has collected 29 million genome-wide SNPs, which have been utilized to map various traits such as Fe and Zn content in grain, tolerance to drought yield, nitrogen-use efficiency, etc. [114]. Genetic loci associated with nitrogen responsiveness and yield traits have recently been identified in foxtail millet through GWAS [115]. Further, integrated breeding associated with genomics facilitated the identification of genetic factors related to seed protein content (SPC), grain yield (GY), and days to maturity (DM) in finger millet [107]. Similarly, genetic regulation of micronutrient content was studied by genotyping-by-sequencing (GBS) and GWAS to identify candidate genes associated with traits in finger millet [116].

Fig. 3.

Overview of recent trends in trait-improvement programs of small millets. Detailed figure is provided as Supplementary Fig. S2.

Fig. 4.

Whole genome sequence data of small millets. Bar diagram showing the number of millet genomes sequenced compared to other major cereals. The data is retrieved from the NCBI Genome database (https://www.ncbi.nlm.nih.gov/genome/).

Table 4.

Summary of genetic transformation efforts performed in small millets for improving their yield-contributing, agronomic, and climate-resilient traits.

Crop	Promoter/Gene	Promoter/Marker system	Vector	Transformation procedure	Phenotype achieved	Reference
Finger millet	CaMV35S/mtlD CaMV35S/UidA	CaMV35S/hpt	pCAMBIA 1380, pCAMBIA 1301	Agrobacterium-mediated	Multiple stress tolerance	[66]
	CaMV35S/SbVPPase	CaMV35S/hpt	pCAMBIA1301	Agrobacterium-mediated	Salt tolerance	[147]
	Ubi1/chi11	CaMV35S/hpt	pCAMBIA1301	Agrobacterium-mediated	resistance to leaf blast disease	[142]
	CaMV35S/gus	CaMV35S/hpt	pCAMBIA1301	Agrobacterium-mediated	–	[141]
	Actin-1/PcSRP	250 mM NaCl	TG0063 of pCAMBIA series	Biolistic	Salt tolerance	[197]
	CaMV35S/PIN	CaMV35S/pac	pPUR	Biolistic	Leaf blast resistance	[144]
Foxtail millet	35S::SiFPGS2/Col-0	CaMV35S	pART-CAM	Agrobacterium-mediated	Increased folate content	[231]
	CaMV45S/gus	CaMV35S/nptII CaMV35S/hpt	pCAMBIA2300, pSB	Agrobacterium-mediated		[21]
	CaMV 35S/SiARDP	CaMV 35S/hpt	pS1300/pCoU	Agrobacterium-mediated	Abiotic stress tolerance	[180]
	CaMV35S/SiLTP	pCAMBIA2300/nptII	pCAMBIA2300/pCOU	Agrobacterium-mediated	Abiotic stress tolerance	[198]
	CaMV35S/SiPHT1;1,2 and 3	pFGC1008/nptII	pFGC1008	Agrobacterium-mediated	High affinity Pi transporters	[199]
	CaMV35S/SiASR4	pCAMBIA2300/nptII	pCAMBIA2300/pCOU	Agrobacterium-mediated	Abiotic stress tolerance	[200]
	CaMV35S/SiMYB3; UBI/SiMYB3	pBI121/GUS; pCAMBIA1390/npt ii	pBI121/pCAMBIA1390	Agrobacterium-mediated; vacuum infiltration method	Tolerance to low-nitrogen stress	[175]
	CaMV35S/SiREM6	CaMV35S/hpt	pS1300	Agrobacterium-mediated; vacuum infiltration method	Salt tolerance	[174]
	Ubiquitin/SiMYB19	Ubiquitin/hpt	pCAMBIA1390	Agrobacterium-mediated	Salt and drought tolerance	[176]
	CaMV35S/SiPf40	CaMV35S/hpt	pCAMBIA1301	Agrobacterium-mediated	Plant architecture	[201]
	Ubi/SiMYB56	Ubi/pat	pMWB014	Agrobacterium-mediated	Drought tolerance	[177]
	CaMV35S/SiATG8a	CaMV35S/hpt	pCAMBIA1302	Agrobacterium-mediated	Tolerance to nitrogen starvation and drought	[181]
	Ubi/SiACC-R	–	pCAMBIA3301	Agrobacterium-mediated	Herbicide resistance and increased oil content	[202]
	Ubi/SiWLIM2b	Ubi/pat	pMWB014	Agrobacterium-mediated	Drought tolerance	[203]
	CaMV35S/SiASR4 CaMV35S/SiASR4	CaMV35S/hpt	pSuper1300	Agrobacterium-mediated	Abiotic stress tolerance	[200]
	CaMV35S/SiMADS51	CaMV35S/hpt	pCAMBIA1302; pCAMBIA1305	Agrobacterium-mediated	Drought tolerance	[204]
	CaMV35S/SiCDPK24	CaMV35S/nptII	pBI121	Agrobacterium-mediated	Drought stress	[42]
	CaMV35S/SiNAC110	CaMV35S/nptII	pBI121	Agrobacterium-mediated	Drought and salt tolerance	[205]
	Ubi::DPY1-3FLAG; CaMV35S::SiBZR1-GFP; CaMV35S/DPY1	Ubi/hpt; CaMV35S/hpt	pTCK303; pEarleyGate 103; pCAMBIA1305	Agrobacterium-mediated	Plant architecture	[206]
	U6a/SiBOR1	U6a/hpt	pYLCRISPR-Cas9-MH	Agrobacterium-mediated	Grain yield	[207]
	CaMV35S/SiGRF	CaMV35S/hpt	pCAMBIA1305	Agrobacterium-mediated	Salt tolerance	[208]
	Ubi/SiBRI1	Ubi/hpt	pCAMBIA1305-eGFP	Agrobacterium-mediated	Stress tolerance	[209]
	CaMV35S/SiHAK1	CaMV35S/nptII	pBI121	Agrobacterium-mediated	Salt tolerance	[210]
	CaMV35S/SiPLDα1	CaMV35S/nptII	pBI121	Agrobacterium-mediated	Drought tolerance	[211]
	CaMV35S/SiBZR1; CaMV35S/SiPLT-L1	CaMV35S/hpt; CaMV35S/hpt	pCAMBIA1305/pGWB5	Agrobacterium-mediated	Drought tolerance	[204]
Teff	CaMV35S/PcGA2ox1	CaMV35S/nptII	pGPTV	Agrobacterium-mediated	Semi-dwarfism	[212]

- Data not available.

Different “omics” approaches, including transcriptomics, proteomics, and metabolomics, enable the quantitative and qualitative analysis of candidate genes to unravel the underlying regulatory networks (Fig. 3). RNA-seq analysis of little millet identified drought- and salt-responsive genes, which can be used as potential crop improvement candidates [15]. Similarly, the transcriptome analysis of finger millet reported several genes associated with drought tolerance and nutraceutical properties, which could be used as candidates for breeding programs [11]. Further, de novo transcriptome sequencing of pearl millet identified salt-responsive genes [117]. Another study by Jaiswal et al. (2018) [118] reported transcriptomic signatures associated with drought tolerance. Recently, small RNA sequencing analysis revealed the function of miRNAs under salt stress and identified their potential targets in pearl millet [119]. Rahman et al. [120] identified salinity-responsive candidate genes such as transporters, stress-related TFs, aquaporins, sodium/calcium exchangers, and signal transducers in finger millet through RNA sequencing. EcbZIP17 from finger millet showed increased growth and stress tolerance under abiotic stresses [121]. The study identified a NAC (no apical meristem) TF, EcNAC67, further functionally characterized in rice using Agrobacterium-mediated transformation [122]. Transcriptomics analysis identified 82 genes encoding calcium sensors in developing spikes of finger millet [123]. Recently, a presumptive model for calcium transport, known as the ‘Tripartite model,’ was suggested in finger millet [124]. Proteomics is another popular omics approach, which is now widely employed in millets to understand climate resilience and nutrition-associated traits. For example, comparative proteomics analysis identified drought-responsive proteins in foxtail millet [125]. Integrating large-scale data from all the omics approaches would facilitate the identification of potential candidate genes that could be manipulated using novel gene-editing tools to develop improved crops.

Genomic studies in small millets have identified several potential markers associated with disease resistance. Field screening has been widely used for screening disease resistance germplasms in small millet [82]. However, these approaches are time-consuming and demand novel phenomics and computational approaches to screening resistant germplasms. Babu et al. [126] identified 58 SSR (simple sequence repeats) markers associated with blast resistance in finger millet. Recently, a genome-wide association study in foxtail millet identified markers associated with blast resistance [127]. Further, QTLs linked with blast resistance in finger millet were identified through association mapping [128]. Functional molecular markers-based resistance genes (R-genes) analogues were identified in finger millet [129]. However, the lack of a complete genome sequence in most small millets hinders the identification of genetic and genomic resources related to disease resistance. Therefore, new-generation biotechnological tools and computational approaches have great potential to accelerate such studies and crop improvement programs in small millets.

8. Next-generation tools for trait improvement in small millets

The improvement of small millets could be a milestone in the “New Green Revolution”- a terminology coined to contemplate novel strategies of crop improvement which are necessary to combat the complex challenges of climate change and malnutrition. In the last few decades, attempts have been made to improve the traits and yield of millets. For example, the comprehensive evolution of pearl millet breeding from an open-pollinated to a hybrid approach has led to approximately a 4% enhancement in pearl millet yield [130]. Further, mutation-breeding in kodo millet has developed non-lodging cultivars, as lodging is a major constraint in its cultivation [131].

Classical breeding programs such as traditional, mutation, and transgenic breeding are very laborious and challenging (Fig. 3). Therefore, new breeding techniques assisted by gene editing (GE), epigenetic modification, and heritable targeted mutation must be applied for crop improvement. Advanced omics have provided substantial genetic and genomic resources, which can be exploited for genetic manipulation using third-generation gnome-editing tools [132]. Clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated protein 9 (Cas9) nuclease system is a powerful gene-editing tool, which is now successfully employed in various crops, including rice (Oryza sativa [133], wheat (Triticum aestivum [134]) and maize (Zea mays [132]). However, this gene-editing tool has not been used for most small millet crops except foxtail millet [135] (Table 5). In this direction, a foxtail millet mutant, xiaomi, with a short generation time, was developed using the CRISPR system [32]. This mutant provides an efficient C₄ model system, which can be utilized to study genes associated with important agronomic traits. Recently, a nano-particle based delivery system for plasmids, ribonucleoproteins (RNPs), and RNA is now developed for speedy trait improvement. However, this concept has not been utilized in millets, but this approach holds a promising future in these crops.

Table 5.

Summary of genome-editing experiments performed in foxtail millet.

Genome editing approach	Transformation procedure	Target gene	Trait	Phenotype achieved	Reference
CRISPR-Cas9	Protoplasts	SiPDS	Carotenoid biosynthesis pathway	Application of protoplast technology	[213]
	Agrobacterium-mediated	SiMTL	Pollen‐specific phospholipase	Haploid inducer lines	[138]
	Agrobacterium-mediated	SiPHYC	Light receptor for photoperiod flowering	Early flowering (heading date of 39 days)	[32]

The genetic manipulation of QTLs is challenging, though GE tools showed promising potential in QTLs editing to instigate desired alleles into several crops by eluding the requirement of excessive crossing [136]. For example, the CRISPR/Cas9 nuclease system facilitated the analysis of phenotypic variation generated by mutations in the cis-regulatory region in tomato [137]. Therefore, the CRISPR strategy could be employed to genetically manipulate multiple QTLs to produce desired phenotypic changes. Further, the generation of double haploids (DH) drastically reduces the generation times, thereby speeding up breeding [136]. Given this, the haploid inducer line of foxtail millet has been generated by CRISPR-Cas9 mediated manipulation of the SiMTL gene [138]. These doubled haploid lines provide a platform to exploit complex traits, including climate resilience, in C₄ crops. This gene-editing assisted de novo domestication of polygenic traits such as climate resilience, and nutritional aspect is a new-generation breeding strategy for crops like small millets.

Developing an efficient transformation and regeneration method is necessary for the successful generation of transgenics by gene editing. Though millets lag on efficient transformation, few millet transformation reports are available [139,140]. For example, Agrobacterium-mediated transformation and regeneration of transgenics have been performed in finger millet [[141], [142], [143]]. Latha et al. (2005) [144] reported the first transgenic finger millet harbouring prawn pin gene encoding PIN fungicide protein, which conferred resistance in response to Pyricularia grisea. Similarly, transgenic finger millet expressing rice chitinase (chi11) was developed through Agrobacterium-mediated transformation, which showed resistance against leaf blast disease [142]. Few other studies reported the development of transgenic finger millet showing salt and drought tolerance through expressing serine-rich protein (PcSrp) and mannitol-1-phosphate dehydrogenase (mtlD) [66,145]. Ramegowda et al. [146] reported the successful development of transgenic finger millet plants, f35S and fBx17, expressing OsZIP1, which showed significantly enhanced Zn and Mn accumulation in finger millet. Following this, a salt-tolerant transgenic finger millet overexpressing the vacuolar pyrophosphate (SbVPPase) gene from Sorghum bicolor was developed through Agrobacterium-mediated transformation [147]. Satish et al. (2017) [143] developed an improved Agrobacterium-mediated transformation for finger millet, which showed better regeneration in four finger millet cultivars. Further, several reports of foxtail transformation are also available [148,149]. Successful transformation using the biolistic method is reported in pearl millet [149]. Similarly, the biolistic method-based transformation has also been tested in barnyard millet [150]. Recently, Agrobacterium-mediated transformation has been reported in kodo millet [151]. Therefore, these successful attempts could set a platform to extend these studies to other small millet crops, accelerating their improvement.

Understanding the genetic diversity of cultivated lines is a crucial part of crop improvement. Resequencing of cultivated lines of several crops such as maize, soybean, and rice has been performed to explore the genetic diversity in terms of SNPs and small insertions/deletions (InDels) that are utilized as markers in genomics-assisted breeding (GAB) [152,153]. Given this, the concept of pangenome has been introduced to capture comprehensive genetic diversities in a species. Recently, the super-pangenome approach has been developed, which means pangenome analysis of pangenomes of various species to explore genetic diversity at the genus level [154]. Despite its significant application in crop improvement, crop pangenomics has not been utilized with small millets. Therefore, pangenomics of wild varieties of small millets could help explore the complete genetic diversity of a genus level, accelerating the small millet improvement programs.

9. Biofortification of small millets for nutritional security

Malnutrition is a major threat in the developing world, where cereals are the primary source of nutrients. Besides major cereals, millets are the major crops cultivated and consumed in semi-arid areas of developing countries. The superior nutritional qualities of millets proved their potential to combat hidden hunger. However, the presence of antinutrients restrains the bioavailability of essential nutrients after consumption. Till now, the biofortification of major cereals has been widely performed to enhance the nutritional qualities of these crops [86]. Subsequently, the biofortification of millets was also initiated to achieve enhanced accumulation and bioavailability of nutrients in millet grains [86]. At present, germplasm conservation for most small millets has been performed in various countries where India has maximum accessions [34]. However, the lack of germplasm functional characterization for nutrients-associated traits limits the biofortification attempts.

Millets are a rich source of carbohydrates, and grains with 0% amylose are highly recommended for infants due to their easy digestibility [86]. However, a lack of information about molecular markers associated with the waxy gene in millets hampers the breeding programs to develop waxy mutants. Therefore, next-generation sequencing approaches could provide the genetic structure of waxy genes in millets [155]. Further, novel gene-editing techniques can generate targeted mutations in non-waxy cultivars for developing waxy cultivars. In this regard, molecular analysis of proso millet, foxtail millet, and barnyard millet has identified mutations in alleles of waxy genes [86,156].

The deficiency of essential amino acids in cereal proteins leads to malnutrition. Millets are a rich source of essential amino acids, where finger millet has the highest content [157]. Therefore, finger millet has excellent potential to be utilized as a model crop to understand the genetic mechanism of protein quality. Kemper et al. [158] identified a gene, o2 modifier (Opm), involved in the modulation of amino acid catabolism, which results in the accumulation of free lysine and tryptophan in the endosperm. Hence, molecular characterization of Opm genes using new advanced biotech approaches will lead to genetic improvement of other cereals and small millets. Further, advanced computational approaches have identified 16 prolamin-encoding genes in foxtail millet, which could be used as a candidate to enhance protein quality in other small millets [159]. Further, the accumulation of micronutrients such as zinc and iron can be improved in cereals and small millets by overexpressing zinc and iron transporters [146]. Similarly, calcium deficiency can be improved as advanced sequencing approaches have identified several calcium sensor genes in small millets [160]. Therefore, the expression and engineering of these calcium sensor genes will help develop calcium-fortified crops. As millets have great synteny with cereals, comparative genomics analysis with available genome sequence information of foxtail millet and other cereals facilitates the identification of orthologous genes associated with several important traits [34].

10. Future of machine learning (ML) approaches for small millets improvement

The modern scenario demands the optimization of large-scale data generated from numerous omics approaches in the agricultural sector. The big data produced by various techniques are difficult to manage because of their “5-V” requirements; 1) Velocity, 2) Volume, 3) Variety, 4) Veracity, and 5) Value [161]. Earlier, conventional statistical methods were extensively used to analyze the genetic diversity, genotypes of crops, yield components, climate resilience, the impact of biotic stress, parental combinations in hybrid breeding, and in vitro biotechnological approaches [162]. However, these conventional methods have low efficacy, as the large-scale data from genomics, transcriptomics, proteomics, metabolomics, and phenomics are non-deterministic and non-linear [162]. Machine learning, a subset of artificial intelligence, has a significant advantage over conventional methods. It can precisely differentiate the plant genotypes based on phenotypical and molecular markers and predict the critical quantitative traits for optimization in vitro breeding methods [162] (Fig. 5). The newly developed phenomics era is also greatly supported by machine learning algorithms. Different sets of algorithms are used to analyze nonlinear data from plant studies. For example, deep learning convolutional neural networks (CNN) facilitate automated phenotyping and disease assessment in plants [163]. Production and yield forecasting is another application of machine learning. Given this, artificial neural networks (ANN) of machine learning showed better performance than classical methods in the production forecasting of pearl millet in Karnataka [164]. Recently, the “Automatic and Intelligent Data Collector and Classifier” was developed by integrating IoT (Internet of Things) and the deep learning approach of machine learning to automatically gather the images and parametric data from rust and blast-infected pearl millet crops [165]. Similarly, a machine learning algorithm, deep learning CNN, was used to identify mildew disease in pearl millet [166]. Another quality testing system, “Mixed Cropping Seed Classifier and Quality Tester (MCSCQT)”, was developed to classify diseased and normal pearl millet and maize seeds [167]. Further, an artificial neural network (ANN) machine learning algorithm was used to develop a model to help farmers predict the suitable crop during the cropping season by using soil conditions and climatic parameters as input [168]. Interestingly, machine learning approaches have applications in predicting the nutrient use efficiencies of crops under field conditions [169]. The major application of machine learning in agriculture has been dedicated to disease and weed detection; however, its implementation in small millet cropping is still lacking. Therefore, a great scope of application of these novel computational techniques in small millet improvement exists.

Fig. 5.

Applications of machine learning in trait improvement. Overview of diverse applications of machine learning in improving the key traits in small millets is shown.

11. Success stories and future prospective of millet omics

Small millets ensure food sustainability due to their inherent climate-resilient features and nutritional superiority compared to rice and wheat. Therefore, crop diversification is essential to attain food and nutritional security in the current scenario. However, low yield, lack of genomic and genetic resources, and presence of several anti-nutrients demand the deployment of advanced biotechnology-based crop improvement approaches to enhance the cultivation of small millets. Owing to this, significant progress has been made in generating enormous genomic and genetic resources for these crops, which led to identifying several agronomically important traits that can be used for small millet improvement.

Conventional breeding approaches have developed several cultivars of small millet with improved climate-resilience features, disease resistance, nutritional quality, yield, improved biomass, and stover quality [[170], [171], [172], [173]]. Currently, 248 varieties of six small millets (finger millet, foxtail millet, proso millet, kodo millet, barnyard millet, and little millet) in India and 19 proso millet in the USA have been released through landraces selection, pedigree selection, and mutation breeding [34]. Interestingly, several stress-tolerance and nutritional quality-related genes identified from small millets have been expressed in other crops, suggesting the applicability of their genetic resources in other crops. For example, the remorin encoding gene (SiREM6) from foxtail millet showed salt tolerance in Arabidopsis on overexpression [174]. Recently, overexpression of SiMYB3 in rice and Arabidopsis conferred tolerance to low nitrogen stress by improving the root architecture [175]. Similarly, overexpression of SiMYB19 enhances yield and salt tolerance in transgenic rice [176]. SiMYB56 overexpression in rice conferred drought tolerance at the reproductive and vegetative stages [177]. Overexpression of EcDehydrin7 from finger millet conferred drought tolerance in tobacco [178]. Late embryogenic abundant (LEA) protein-encoding gene, SiLEA14, from foxtail millet, conferred salt and osmotic stress tolerance in transgenic Arabidopsis and foxtail millet [21]. Numerous other studies have reported the functional characterization of important genes from small millets in other crops [[179], [180], [181], [182], [183], [211], [214], [215], [216], [217], [218], [219], [220], [221], [222], [223], [224], [225], [226], [227], [228], [229], [230]]. However, such efforts are lacking in small millets due to the absence of standardized transformation methods. Further, despite the availability of advanced breeding approaches, very limited trait improvement efforts have been made in most small millets, including fonio, guinea millet, browntop millet, and job's tears. This is mainly due to meagre germplasm conservation for these small millets, which restrains crop improvement efforts. However, the high-quality cross-transferability of markers demonstrates the potential scope of marker-assisted breeding in small millets crops with limited genomic resources. Conventional breeding programs have shown significant success in delivering improved cultivars of small millets, though many have to be explored in these crops. The advanced sequencing approaches could be employed to provide a complete genome sequence of several small millets, including kodo millet, job's tears, and little millet. Additionally, novel machine learning-assisted phenomics approaches could further accelerate the collection of genetic diversity data. Given the current looming climatic conditions and increasing malnutrition, the agricultural community has increased its focus on small millet improvement to achieve food, nutrition, and economic security.

Author contribution statement

All authors listed have significantly contributed to the development and the writing of this article.

Funding statement

Dr. Mehanathan Muthamilarasan was supported by the Institute of Eminence Grant awarded to the University of Hyderabad by the Ministry of Education, India [UoH-IoE-RC2-21-014].

Data availability statement

No data was used for the research described in the article.

Declaration of interest's statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the supplementary data related to this article: