Peanut (Arachis hypogaea L.) widely known as groundnut is a cultivated allotetraploid (2n = 4x = 40), particularly valued for its protein content (28%), oil content (50%) and is rich source of calcium, phosphorus, potassium, magnesium, iron, zinc, and boron. The peanuts also contain vitamin E and small amounts of vitamin B complex (1). Due to these properties, it is rightly called as “king of oilseeds” and “poor man's almond” and provides 567 kcal of energy per 100 g of kernels. Peanut is derived from a natural hybridization process between A. duranensis (AA genome progenitor) and A. ipaensis (BB genome progenitor) resulted in amphidiploid A. monticola. Later in the evolutionary process fertile, now cultivated allotetraploid A. hypogaea was evolved. The genome sequence of cultivated peanut revealed a genome size of 2.7 GB (2, 3). The global production of peanut is estimated to be 47.09 million tons from 27.94 million hectare area with productivity of 1685.6 kg ha⁻¹ (4). More than 50% of the peanut produced are used for oil extraction. Rest are either consumed as kernels or processed into various products like peanut butter, milk, cheese analogs, beverages, plumpy nut (a ready to use therapeutic food), and chocolate additives. Peanut is considered as both legume (botanically) and nut (nutritionally). As a plant based protein, it also contains wholesome amount of fiber and other bioactive functional compounds that improve health in humans including heart health, weight loss, gallstone prevention, malnutrition, etc. (5–9). For instance, “Plumpy nuts,” a ready-to-use therapeutic food (RUFT) is being used for the treatment of severe acute malnutrition (SAM) which affect adversely about 20 million children globally, with approximately 8 million cases in India alone. RUTF use is very common in Africa and other developing countries where the incidence of SAM is high. Peanut has several functional components with variety of health benefits like coenzyme Q10 which helps the heart protection under low oxygen areas like high altitude and clogged arteries, rich arginine helps in blood circulation, resveratrol protects against cancer, cognitive disorder and Alzheimer's disease, phytosterols of peanut butter, oil reduces the absorption of cholesterol from the diet, magnesium and dietary fiber reduce the diabetes, reduces LDL cholesterol and helps in weight management [reviewed by (5)]. The foods with different colors including fruits, vegetables, legumes and cereals are found to have functional components that are beneficial for the human health (10). Peanuts have different tinge of kernel colors varying from the white to dark purple with solid or variegated color patterns. The peanut germplasm includes peanut cultivars with varying kernel colors. For selection of appropriate lines that retain the diversity persisting in the global collection becomes essential in utilizing in crop breeding programs. In this regard, the “core collection” (i.e., 10% of the entire germplasm collection) presents a manageable and cost-effective entry point into germplasm collections for identifying candidate genotypes for various traits (11). In peanut, the core collection were further narrowed down to mini core collections and U. S. peanut mini core collection (12), ICRISAT mini core collection (13), and Chinese peanut mini core collection (14). Traits related to abiotic stresses (drought, heat, salinity, low temperature, and P deficiency, calcium induced iron chlorosis), biotic stresses (early leaf spot, late leaf spot, rust, tobacco cut worm, Aspergillus flavus, peanut bud necrosis disease, and bacterial wilt) and seed quality (oil, protein, oleic/linoleic ratio, Fe and Zn) have been screened by many researchers in ICRISAT mini core collection (15–17). GWAS for 50 different agronomic traits was carried out in reference set of core collection having 300 genotypes which included the 184 mini core genotypes indicated several markers associated with major agronomic traits (18). The efforts toward elucidating the marker-trait association related to quality or nutritional parameters are limited to some crops like sorghum (19), wheat (20), rice, and barley (21). Evidence of metabotype-phenotype linkage was explained by using parallel metabolite and phenotypic GWAS in traits such as grain color and size in rice and maize kernels (22).

Estimation of proximates like moisture content (MC), fat, crude protein (CP), ash, crude fiber (CF), and carbohydrates (CHO) provide the distribution of major components in any food or processed product. The peanut is rich in nutraceuticals like total polyphenol content (TPC) and total antioxidant activity (AOA). Proximates and nutraceuticals content in the kernel determine the quality of peanut. Peanuts must be dried or cured to guarantee that, the moisture content does not surpass 10.5%, to guarantee quality and to avoid the development of microorganisms. Low moisture percentage of peanut seed prevents it from the susceptibility to the aflatoxin producing fungal pathogens like Aspergillus flavus (23). Fatty acid composition in peanut is heart-friendly. Oleic acid (monounsaturated omega-9 fatty acid) is important seed quality parameter and has inverse association with systolic blood pressure. Digestibility of peanut protein is similar to that of animal protein (24). Crude fiber in peanut has little food value but provide the proper help in the intestinal tract for adequate peristaltic action. It contains low glycemic index (GI) and glycemic load (GL) and addition of peanut butter to high glycemic load food can stabilize the blood glucose level of the body. Nutraceuticals of peanut gives health advantages, including the aversion or potential treatment for the diseases [reviewed by (5)].

Peanut pod is made up of external shell (21–29%), kernel (69–73%), germ (2–3.5%), and testa or seed coat (2–3%). Due to papery like structure, seed coat is also called as peanut skin or peanut kernel skin (PKS). Several studies showed that PKS are rich in polyphenol and antioxidant [reviewed by (25)]. The kernel skin is removed through blanching or roasting before the preparation of snack food, groundnut butter and other groundnut-based foods and PKS remained as by-product. High fiber and tannin content of PKS make them usable only for ruminants (and possibly rabbits) but, due to chance of aflatoxin contamination, utilization of PKS as feed remains limited with inclusion rates lower than 10% of the diet as feed for ruminants (26). The reason for the variation in skin color was reported to be due to flavonoid content in cereals, legumes, oilseeds, and several other colored foods (27–29). Due to growing evidence of the versatile health benefits of dietary flavonoids through epidemiological studies (30), the PKS emerges as the potential product to be utilized in diet.

Peanut skin color was found to be strongly associated with total phenolic content and hue angle was proposed to be a biomarker for total polyphenol content and antioxidant capacity when 27 cultivars were screened (31). Similarly total anthocyanin content in the peanut skin was found to be associated with the skin color when 22 selected genotypes from US mini core collection along with 4 Israeli Virginia type cultivars were screened (32). The relation between TPC, flavonoids and AOA was studied with grain color in 481 accessions of rice with white, red and black colored genotypes and correlation studies indicated a negative correlation between a* and antioxidant capacity (33). The grain color was also found to be associated with anthocyanin and Zn content when 156 rice accessions varying in grain color were screened and analyzed through genome-wide association studies (34). Similar findings were revealed from transcriptomics and proteomics studies in colored rice (22). Similarly, GWAS studies have been carried out to dissect the candidate genomic regions for flour color using colorimetric approaches (35).

Different sized peanuts are preferred for different purposes. For instance, peanuts with large seed, low oil but with high oleic acid/linoleic acid (O/L) ratio are preferred for direct consumption, while medium seeded peanuts are preferred by the industries. Studies related to kernel size are limited to inheritance studies (36). A study indicated that kernel size was not significantly correlated with the oil content. Protein content was higher in small seeded compared to the bold seeded peanuts (37). However, in soybean no significant correlation was found between protein and oil content with seed size (38). In peanut the studies on kernel color are limited to the estimation of TPC and flavonoids through biochemical assays and correlation studies thereof. There were limited efforts to identify the genomic regions underlying these traits. In this regard, the molecular markers are considered to be powerful genomic tools to characterize the genetic variation present in the population. The molecular markers can be used in trait mapping and molecular breeding programs [reviewed in (39)]. Low level of genetic variation was observed in peanut. Transposable elements (TE) provide an important source of variation and are highly dynamic in diverse species. Due to the positive correlation of transposable element content with the genome size of the organism, it has been widely known as a genomic parasite and found to be source of variation in plants and animals (40). Use of DNA transposon markers in peanut was proposed by Bhat et al. (41) and presence or absence of AhMITE1 at a predetermined site in the genome was confirmed using PCR (42). There are two types of transposable elements viz., class I TE (move within genomes via RNA intermediates, using a copy-and-paste mechanism) and class II TE (DNA of a DNA transposon moves by a cut-and-paste mechanism). Class II TEs entail autonomous and non-autonomous elements. Miniature inverted-repeat transposable elements (MITEs) are non-autonomous elements of less than 600 bp in length (43). It contains 10-15 bp terminal inverted repeats and two flanking target site duplications. MITEs are inserted mainly in the gene-rich regions and can affect gene expression (44). In peanut, due to the availability of genome sequence information, several TE markers commonly known as Arachis hypogaea transposable element (AhTE) have been identified and are made available in a database at Kazusa Peanut Database (45). AhTE markers have shown greater potential to differentiate the genotypes in groundnut (42, 43, 46). As AhTE markers can be easily screened on agarose gel through electrophoresis, they have been utilized in trait mapping using linkage and association mapping approaches (41–43, 47–51). The candidate genes or genomic regions governing the nutritional traits are presently not available. Hence an effort was made to identify genomic regions that are associated with the nutritional traits including proximates and nutraceuticals like TPC and AOA genotypes that vary in kernel skin color and size.

Genetic improvement of peanut for quality and yield traits is essential for combating the malnutrition and hunger in the developing countries. Major focus of the peanut breeding programs hovers around increasing the yield potential of the crop to increase the productivity under biotic and abiotic stresses. Advances in genomics technologies in peanut through next generation sequencing, have deciphered tremendous genomic resources that can be utilized in genomics-assisted breeding. In this regard, efforts were made to identify molecular markers or genomic regions associated with traits such as late leaf spot and rust (57–61), aflatoxin contamination (62), drought (63), yield related traits (50), and some quality traits like protein content, oil content, and oil quality (64–66). Another important trait that needs attention in peanut is its skin color- more commonly called as peanut kernel skin (PKS) color. The association of skin color with various nutritional parameters is well established in various vegetables and fruits (30). In food crops, the relationship between flavonoids and antioxidant activity was studied with grain color in colored rice (33, 34). However, in peanut the studies are limited to studying the relation between PKS with nutraceuticals like total polyphenol content and antioxidant capacity (31, 32). In this context this, present study reports the profiling of nutritional parameters in sixty genotypes of peanut that differ in skin color and size.

Colorimetric method was used for quantification of skin color that had estimated 11 color parameters. For color estimation, there are many color models like munsell color space, RGB/CMYK, YIQ/YUV/YCbCr, HIS/HSV/HSL, Hunter Lab color space, and CIEXYZ/CIEL*U*V*/CIEL*a*b*. Among these models CIE model and Hunter Lab color space are commonly used for evaluation of color difference while other have their application area in computer graphics, image analysis, and processing, etc. (67, 68). Both the color space are based on opponent-colors theory (69). In the present study CIELAB values were derived by using the colorimetric approach to quantify the peanut skin color. Similar approach was used to study the skin color in 27 peanut cultivars (31), 481 colored rice accessions (33). Similarly hunter scale was used for quantification of color estimation of skin color of 22 US peanut mini core collection and 4 Israeli cultivars (32), 17 colored chickpea lines (70), and 33 cool season legumes (71). These color parameters were also considered as the biomarkers for estimating the TPC and AOA in peanut skin (31). Hence color strength was used for clustering the 60 genotypes used in the study by DMRT analysis and 10 distinct groups were made. Cluster-I consisted of dark purple while cluster-X with white colored genotypes. Cluster-X with white colored genotypes showed the lowest fat content, TPC, and AOA indicating light skin colored genotypes were poor in the nutraceuticals while dark colored seeds are rich in nutraceuticals compared to white seeded genotypes. Similar observations were recorded in peanut (31, 32), rice (33), chickpea (70). In this study light pink colored seeds showed highest antioxidant activity rather than expected dark purple colored seeds, which may be due to colorless flavonoid that exhibit strong antioxidant activity. Similar finding i.e. high TPC was observed in light pink colored peanut kernel skin (32). Cluster-X with white skin colored genotypes showed highest protein content and lowest fat content, further the study also showed the negative correlation between fat content and protein content. This clearly suggests that lighter skin color is an indicator of high protein and low fat content. The flavonoids were found to have inhibitory action on protein production in eukaryotes (72, 73). In the present study peanut skin was used to assess the correlation with nutritional parameters in the whole kernel (including skin), that provides better understanding of the nutritional status of the peanut seeds. There are some reports where nutraceuticals were measured in peanut skin devoid of kernel (32). Presently, there are no published reports to correlate proximates with skin color in peanut; hence this study will provide the prime information on proximate profiling. As expected, white colored genotypes showed lowest reflectance value and highest moisture content. The moisture content was lowest in dark tan (Cluster-VII) followed by dark and light purple colored (Cluster-I and II) seeds which contain lower moisture content maybe due to their high light absorbance nature. Generally, colored/darker grains have higher nutritional value compared to their light counterparts. For instance, the dark colored rice were nutrient rich compared white rice (74). In this study also colored seeds showed higher nutritional value compared to white and light colored seeds except crude protein content. The AOA in the present study was comparatively higher in colored seeds than the white ones; this kind of observation was also evident in rice (33). The TPC content was found to be highest in dark red (Cluster-III) colored seeds followed by light red (Cluster-IV) and lowest in white seeds (Cluster-X). But, the findings from other peanut skin studies found that light pink colored skins followed by red had larger amount of TPC content (32). The crude fiber content was maximum in light red (Cluster-IV) and light tan (Cluster-IX) and minimum in dark red and dark pink seeds (Cluster III and V).

The inheritance studies of peanut skin color carried out way back by Higgins (75) shows that the flesh-colored seed testa is dominant over white color with bigenic difference, whereas the red testa color is dominant to flesh-colored testa with monogenic difference. It also showed the necessity of flesh pigment for the expression of red color. These results were further confirmed on different white colored peanuts by Hammons (76). However, Norden et al. (77) revealed the presence of single dominant gene that was epistemic to the previously described genes for testa color in peanut. A contradictory observation where, white testa color was found to be dominant in one of the genotypes (78). Further, three different alleles were identified for the red testa color and among them; R1 allele was to be dominant over the other two (r2 and r3) recessive alleles (79). In another study, the purple and purple striped- testa in peanut showed monogenic and bigenic inheritance respectively (80). The observations made on the inheritance studies in peanut showed debatable results, indicating the complex nature of seed coat color. Since then, there are limited efforts to elucidate the mechanisms involved in seed coat color. The present study provides some insight into seed coat color and its association with AhTE markers like, AhTE1581-color strength, AhTE0189-reddness-greenness of the color, AhTE2011-ligthness-darkness of the color, AhTE1777-hue angle of the kernel skin color. Further investigations are necessary to understand the molecular mechanisms underlying testa color determination using holistic genomics and metabolomics approaches.

Generally, Virginia peanut cultivars are bold seeded, majority of Valencia are medium sized and Spanish types are small seeded genotypes. In the present study based on test weight genotypes were classified into small, medium, and bold seeded types. The kernel size had no effect on any of the nutritional parameters except for crude protein according to the correlation studies. A positive correlation of kernel size with protein content was observed which was contrary to the previous observation in peanut by Prathiba and Reddy (37). However, no significant correlations between seed size with protein content was observed in soybean (38). Hence the relation between kernel size and protein content is debatable and needs further investigation.

Correlation studies showed the negative association of fat content and protein which was in accordance with the findings by Dwivedi et al. (81). Number of pods per plant showed negative correlation with plant height and positive correlation with pod weight per plant. As the plant height increases, the peg penetrating the soil to make pod decreases. Hence the number of pods per plant as well as pod weight per plant decreases (82). In the present study, number of pods per plant and pod weight per plant showed high GCV coupled with PCV which was also in accordance with the findings of Hake et al. (50).

The studies on seed color and their relation with other nutrients in peanut and many other crops was restricted to the biochemical analysis and lacked the molecular approach to identify probable candidate/ genomic regions. Hence for molecular characterization, AhTE markers that show higher polymorphism were utilized in this study with PIC value ranging from 0.03 to 0.37. The heterozygosity ranged from 0.03 to 0.47, with mean 0.23 across the genotypes. This may be due to the biallelic nature of AhTE markers and also the type of population which consist of diverse genotypes. This finding was in accordance with previous peanut studies (50, 51). About 41 out of 100 AhTE markers screened across 60 genotypes were found to be polymorphic. The genotypic data was used to study the structure of the population. The study revealed that there were two subpopulations that were grouped based on the botanical types. Though variability was found with respect to phenotype, genotypically the population showed less variation. Population structure was analyzed in different populations in peanut (like mini core collections, reference collections, mutant populations) and found that the grouping was made as per the botanical varieties of Arachis hypogaea, which include hypogaea, hirsuta, fastigiata, vulgaris etc. (18, 50, 83).

Most of these traits are quantitative in nature and are influenced by number of minor genes. The single marker analysis showed about 110 MTAs with 75 major and 35 minor MTAs. Out of which, 43 MTAs were detected for 9 morphological and yield related traits (leaflet length, leaflet width, plant height, number of pods per plant, pod weight per plant, shelling percentage, test weight, sound mature kernel weight percentage, and haulm weight). AhTE0465 (leaflet length, leaflet width, and shelling percentage), AhTE1253 (pod weight per plant and test weight), and AhTE1907 (plant height, shelling percentage, test weight and haulm weight) showed association with only morphological and yield related traits. The marker AhTE0205 on chromosome number A03 showed association with plant height. In another study, a SNP marker with location A03-26481539 was found to be associated with plant height in peanut using GWAS approach (84).

Few MTAs were also observed for color parameters like AhTE1581 (K/S and RFL), AhTE0189 (a*), AhTE1777 (hue angle and Δa), AhTE2011 (L*), and localized on chromosomes B01, B04, and B08. A genome wide association study in peanut also indicated a SNP with location B03-22076736 position had significant association with seed coat color (44). The occurrence of associated markers on B genome might indicate the genome dominance for color parameter that needs to be further investigated.

Out of 75 major MTAs, 24 were related to nutritional traits. Markers associated with AOA include AhTE0233, AhTE0205, AhTE0474, AhTE1542, AhTE0189, AhTE0446, AhTE0143, and highest PVE (18.25%) was explained for AhTE0474 which is localized on chromosome A03. Markers associated with TPC include AhTE0442, AhTE0189, AhTE0362, AhTE0143 and highest PVE (24.56%) was explained for AhTE0189. AhTE0189 and AhTE0143 were fund to be associated with both the nutraceuticals i.e., TPC and AOA. Trait mapping in sorghum through GWAS found candidate genes that are associated with TPC (19). Except for moisture content, markers associated with all the proximates were identified in this study. For example, AhTE2000 marker localized at chromosome B08 was associated with fat content, crude protein and crude fiber. Similarly, AhTE1761 on chromosome B04 was found be associated with fat content and crude protein content. In peanut, association studies on quality parameters using molecular markers were limited to oil content, protein, Fe and Zn content and no published records are available for proximates and nutraceuticals. Previous reports on genome-wide association on reference set of peanut consisting of 300 genotypes showed MTAs for quality traits like, oil (25 MTAs), oleic acid (2 MTAs) protein (11 MTAs), oleic/linoleic acid (22 MTAs), and zinc (1 MTA) (18). Also, highly significant 38 MTAs for protein and oil related traits were observed (50).

The associated markers were checked for their location on the peanut genome from PeanutBase. The chromosome B04 had five markers that were found to be associated with a*, AOA, TPC, leaflet length, haulm weight, fat, crude protein, Δa, hue angle, and shelling percentage. Chromosomes A03, A06, and B06 had 3 markers each that were found to be associated with 5 traits (haulm weight, shelling percentage, plant height, AOA, and TPC), 4 traits (carbohydrate, ash, sound mature kernel weight percentage, and leaflet length) and 6 traits (carbohydrate, plant height, shelling percentage, test weight, haulm weight, reflectance), respectively. Further gene annotation for the candidate genes of some of the marker like AhTE0442, AhTE0205, and AhTE0474 on chromosome number A03 showed functions related to oxidoreductase activity. Several reports in peanut showed markers associated with disease resistance were also localized on A03 that was evident from QTL mapping and sequencing (59, 60, 85). The occurrence of genes encoding for antioxidant activity at A03 might provide disease resistance in peanut. However, large number of markers needs to be screened on a larger population to identify the candidate genes/markers using genome-wide association analysis to validate these findings. Further, the associated markers after validation and identified genotypes with favorable alleles can be utilized in molecular breeding for nutritionally rich peanuts.

In the current scenario, improving the peanut crop with quality parameters is equally important along with increasing yield and pest resistance. In order to combat the malnutrition and hunger, we need to work toward “more nutrition per bite.” This study provides relation of peanut skin color with various nutritional parameters (proximates, nutraceuticals), morphological and yield related traits. The markers associated with important traits after validation and genotypes with favorable alleles can be utilized in genomics-assisted peanut improvement.

All datasets generated for this study are included in the article/Supplementary Material.

SN and HN conceived the presented idea. VH performed the experiment. GN provided the seed material. SN, PB, and HN designed the experiments. RB provided the molecular markers. PB facilitated lab for nutrition estimation and helped in conducting nutritional experiments. SN and VH analyzed the data, and wrote the manuscript. SN supervised the project and arranged for funding.