Nội dung

Pradhan et al. BMC Plant Biology (2020) 20:57 https://doi.org/10.1186/s12870-020-2262-4 RESEARCH ARTICLE Open Access Linkage disequilibrium mapping for grain Fe and Zn enhancing QTLs useful for nutrient dense rice breeding S. K. Pradhan1*†, E. Pandit1*†, S. Pawar1, R. Naveenkumar1, S. R. Barik1, S. P. Mohanty1, D. K. Nayak1, S. K. Ghritlahre1, D. Sanjiba Rao2, J. N. Reddy1 and S. S. C. Patnaik1 Abstract Background: High yielding rice varieties are usually low in grain iron (Fe) and zinc (Zn) content. These two micronutrients are involved in many enzymatic activities, lack of which cause many disorders in human body. Biofortification is a cheaper and easier way to improve the content of these nutrients in rice grain. Results: A population panel was prepared representing all the phenotypic classes for grain Fe-Zn content from 485 germplasm lines. The panel was studied for genetic diversity, population structure and association mapping of grain Fe-Zn content in the milled rice. The population showed linkage disequilibrium showing deviation of HardyWeinberg’s expectation for Fe-Zn content in rice. Population structure at K = 3 categorized the panel population into distinct sub-populations corroborating with their grain Fe-Zn content. STRUCTURE analysis revealed a common primary ancestor for each sub-population. Novel quantitative trait loci (QTLs) namely qFe3.3 and qFe7.3 for grain Fe and qZn2.2, qZn8.3 and qZn12.3 for Zn content were detected using association mapping. Four QTLs, namely qFe3.3, qFe7.3, qFe8.1 and qFe12.2 for grain Fe content were detected to be co-localized with qZn3.1, qZn7, qZn8.3 and qZn12.3 QTLs controlling grain Zn content, respectively. Additionally, some Fe-Zn controlling QTLs were co-localized with the yield component QTLs, qTBGW, OsSPL14 and qPN. The QTLs qFe1.1, qFe3.1, qFe5.1, qFe7.1, qFe8.1, qZn6, qZn7 and gRMm9–1 for grain Fe-Zn content reported in earlier studies were validated in this study. Conclusion: Novel QTLs, qFe3.3 and qFe7.3 for grain Fe and qZn2.2, qZn8.3 and qZn12.3 for Zn content were detected for these two traits. Four Fe-Zn controlling QTLs and few yield component QTLs were detected to be colocalized. The QTLs, qFe1.1, qFe3.1, qFe5.1, qFe7.1, qFe8.1, qFe3.3, qFe7.3, qZn6, qZn7, qZn2.2, qZn8.3 and qZn12.3 will be useful for biofortification of the micronutrients. Simultaneous enhancement of Fe-Zn content may be possible with yield component traits in rice. Keywords: Association study, Linkage disequilibrium, Grain Fe content, Grain Zn content, Biofortification Background Majority of the global population consume rice daily, particularly in Asiatic countries. But, rice grain is poor source of micronutrients such as iron (Fe) and zinc (Zn). The practice of consuming polished rice as a staple food in India aggravates malnutrition. Substantial amounts of iron and zinc are removed during milling. The polished rice contains around 2 mg kg− 1 Fe while the recommended * Correspondence: pradhancrri@gmail.com; elsambio@gmail.com † S. K. Pradhan and E. Pandit contributed equally to this work. 1 ICAR-National Rice Research Institute, Cuttack, Odisha, India Full list of author information is available at the end of the article dietary intake is 10–15 mg kg− 1. Similarly, polished rice contains around 12 mg kg− 1 of Zn, whereas the recommended intake for humans is 12–15 mg kg− 1 [1]. Iron is an important constituent of haemoglobin in red blood cells and is essential for the proper functioning of several enzymes in the body. Zinc is required for the metabolic activity of 300 enzymes, and is essential for enzymes involved in cell division, protein synthesis and growth [2]. To date, proper attention has not been given for improvement of these micronutrients in rice grain. Literature survey on existence of grain Fe and Zn content diversity reported in rice is high in natural rice germplasms [3–13]. © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Pradhan et al. BMC Plant Biology (2020) 20:57 Increasing the iron and zinc content in rice grain through breeding is cheaper and an easier option to reduce malnutrition in the developing countries. Increasing Fe and Zn content in rice is possible by utilizing elite germplasms possessing enormous genetic potential for grain Fe-Zn content in rice breeding program [5, 6, 13–15]. Popular rice varieties usually contain lesser micronutrients in grains compared to the traditional cultivars and landraces [16, 17]. QTLs responsible for further enhancing these nutrients in the high yielding rice varieties need to be identified and utilized. Through classical breeding approach, a very limited success could be achieved in this context. Earlier reports on low heritability of micronutritients concentration, negative relationship with high grain yield and genotype and environment interaction were the main limitations of genetic enhancement for these nutrients [18–21]. However, the development of CR Dhan 311, DRR Dhan 45 and Chhattishgarh Zinc Rice-1 are the recent examples of high yielding varieties in India with high grain Zn content. But, these achievements were through non-targeted classical breeding approach. Success in breeding for micronutrients enrichment is limited due to involvement of many quantitative trait loci (QTLs) with small effects and interacts highly with the environmental factors. Precision breeding is a better approach for enriching the elite rice varieties with desired micronutrients. The availability of information on robust markers for different QTLs and potential donors are pre-requisite for the success of precision breeding. Several QTLs for grain Fe and Zn related traits have been reported in rice from different genetic backgrounds of intraspecific and interspecific crosses [6, 9, 13–15, 18–36] are hardly being in use in modern molecular breeding for Fe-Zn enhancement. Mapping reports of Fe-Zn content in rice indicated the use of various bi-parental mapping populations for detecting the genes controlling these two micronutrients. Gene mapping using bi-parental mapping populations is more time consuming, costlier and lesser resolution than association mapping [17, 37–40]. These limitations can be overcome by linkage disequilibrium (LD) mapping or association mapping. Association mapping is based on linkage disequilibrium or variations existing in natural or developed populations [17, 37, 39, 40]. The main purpose of association mapping here is to estimate the correlations between molecular markers with the grain Fe and Zn content in a panel containing representative rice population that possess considerable amount of variability for these two traits. Only few research reports on association mapping for grain Fe-Zn content are available [41, 42]. But, report on QTL mapping through association mapping in rice are available for grain yield and agronomic traits [29, 41, 43, 44]; seedling stage cold tolerance [40, 45]; cold tolerance at germination and booting stages [46]; high temperature stress tolerance [39]; grain quality traits [47]; Page 2 of 24 salinity tolerance [48]; drought tolerance traits [49, 50]; early seedling vigour [51] and grain protein content [17]. However, the information on association mapping for grain Fe and Zn content in rice is scarce. In this study, a panel population containing 102 genotypes shortlisted from 485 germplasm lines were analyzed using 100 SSR and gene specific molecular markers for genetic diversity, population structure and association mapping for grain Fe and Zn content. Results Phenotyping of the panel population for grain Fe and Zn content in milled rice The panel population was prepared by shortlisting genotypes from different phenotypic groups based on screening results of 485 germplasm lines for Fe-Zn content. Almost equal proportions of genotypes were pooled from each Fe-Zn phenotypic group along with the check varieties for constitution of the panel containing 102 germplasm lines. The mean iron content in of the panel population varied from 1.07 to 5.38 mg kg− 1 in the milled rice phenotyping results of wet seasons, 2016 and 2017. A higher Fe-content of > 4 mg kg− 1 in milled rice was observed in 13 genotypes viz., Chittimuthylu, IET25450, IET 24775, IET 24760, IET 24316, BPT5204, IET 23832, Kalanamak, IET25441 and IET25465, IET 23829 and IET 24779 (Additional file 3: Table S1). The check varieties for micronutrients viz., Chittimuthylu and Kalanamak showed mean grain Fe content of 4.8 and 4.48 mg kg− 1 , respectively. Genotypes containing ≥25 mg kg− 1 zinc and /or ≥ 10 mg kg− 1 iron in brown rice along with at par or higher grain yield compare to the yield checks are considered desirable biofortified rice lines. While comparing in milled rice, genotype containing ≥4 mg kg− 1 Fe and ≥ 20 mg kg− 1 Zn and producing higher yield than yield check variety is considered as desirable genotypes. It is observed that the genotypes enlisted in the Table containing > 4 mg kg− 1 Fe in milled rice were found to be with > 12 mg kg− 1 in brown rice. Similarly, genotypes present in the panel with > 20 mg kg− 1 Zn content in milled rice showed > 25 mg kg− 1 in brown rice. From the results, nine biofortified rice genotypes, two micronutrients check cultivars and landrace Sadakadam showed more than 4 mg kg− 1 Fe in milled rice and categorized under high Fe containing genotypes in our study (Table 1). Forty two genotypes comprised of the moderate group (3–4 mg kg− 1 Fe in milled rice) while 48 showed low values (< 3 mg kg− 1 Fe). Mean zinc content in milled rice varied from 7.43–27.97 mg kg− 1 showing ≥20 mg kg− 1 in fourteen genotypes (Additional file 3: Table S1). These fourteen genotypes were categorized under high grain Zn containing rice. Pradhan et al. BMC Plant Biology (2020) 20:57 Page 3 of 24 Table 1 Mean estimate values of days to 50% flowering, grain Fe- Zn content, panicles/m2 and grain yield of 102 genotypes including biofortified lines and check varieties studied during wet season, 2016 & 2017 Sl. No. National testing No./Accession No./Name of the genotype Days to 50% flowering Iron content in ppm Zinc content in ppm Panicles/m2 Grain yield (kg/ha) 1 IET23829 99 4.18 18.73 282 3950 2 IR64 91 3.97 17.24 281 4250 3 IET23834 88 3.5 18.8 264 3450 4 Kalanamak 106 4.48 25.91 265 2962 5 IET23824 90 3.75 20.1 262 3771 6 IET23832 97 4.15 18.43 266 4068 7 Chittimuthyalu 110 4.8 22.03 283 3620 8 IET24780 94 3.21 14.09 295 4552 9 BPT5204 111 4.11 15.46 285 4431 10 IET24771 107 3.9 15.68 286 4958 11 IET24766 94 3.41 15.2 279 4281 12 IET24316 87 4.04 18.25 264 3532 13 IET24391 110 2.99 20.11 249 3984 14 IET24777 96 2.73 14.8 295 4848 15 IET24760 110 4.11 18.88 303 4693 16 IET24775 110 4.33 17.47 284 4623 17 IET24783 96 3.35 17.18 275 4883 18 IET24544 98 2.49 16.4 286 4899 19 IET24336 93 2.77 17.37 269 4598 20 IET24772 96 3.26 17.8 287 4285 21 IET24557 107 3.97 18.34 288 4079 22 IET24779 102 4.02 19.67 280 4283 23 IET24774 104 2.93 17.4 293 4685 24 IET24787 97 3.62 17.2 280 3826 25 IET25441 101 4.06 16.48 275 4258 26 IET25443 109 3.6 19.73 284 3629 27 IET25444 111 3.45 16.31 257 4229 28 IET25445 106 3.57 14.55 287 4399 29 IET25446 94 3.95 21.02 284 3857 30 IET25447 105 3.5 15.57 290 4748 31 IET25449 98 3.19 15.2 266 4379 32 IET25450 99 4.41 23.43 282 3133 33 IET25452 106 3.41 16.42 280 4490 34 IET25453 93 3.03 15.95 269 4515 35 IET25454 87 3.33 19.27 227 2550 36 IET25457 99 2.9 16.71 246 3582 37 IET25459 102 2.97 15.22 272 4479 38 IET25460 105 3.41 14.22 284 5210 39 IET25461 95 3.53 20.14 256 3767 40 IET25463 98 3.09 15.32 276 4843 41 IET25464 94 2.76 16.2 243 3392 42 IET25465 101 4.19 20.36 265 3656 Pradhan et al. BMC Plant Biology (2020) 20:57 Page 4 of 24 Table 1 Mean estimate values of days to 50% flowering, grain Fe- Zn content, panicles/m2 and grain yield of 102 genotypes including biofortified lines and check varieties studied during wet season, 2016 & 2017 (Continued) Sl. No. National testing No./Accession No./Name of the genotype Days to 50% flowering Iron content in ppm Zinc content in ppm Panicles/m2 Grain yield (kg/ha) 43 Gontra Bidhan3 106 2.7 11.1 251 4050 44 IET25469 95 3.15 18.57 276 4054 45 IET25470 96 2.98 18.1 263 4772 46 IET25471 103 3.44 16.87 265 4276 47 DRRH3 101 2.53 13.88 253 5162 48 IET25472 92 3.78 26.96 275 3059 49 IET25473 103 3.35 16.62 280 5002 50 IET25474 105 3.26 18.95 269 4125 51 IET25475 91 3.37 23.35 267 3881 52 IET25477 101 3.13 24.28 284 3620 53 IET25478 105 2.94 16 271 4209 54 IET25479 106 3.94 16.84 290 4424 55 Lalmeeta 118 2.79 12.66 252 3430 56 Abhimanyu 119 3.78 10.44 272 3810 57 Kalobhutia 121 3.96 11.7 246 3360 58 Sadakajam 115 4.17 16.32 246 3210 59 Geetanjali 128 3.18 11.94 280 3950 60 Kakhru 127 2.85 13.56 254 3740 61 Boanti 120 3.18 13.56 248 3430 62 Tulsimukul 120 3.99 14.1 226 2830 63 Kokilpatri 112 2.94 11.52 252 3280 64 Basmatikarnal 125 3.21 12.3 242 3120 65 Kalonunia 132 3.84 13.2 218 2930 66 SafedLuchai2 108 3.69 16.5 220 2940 67 Bankra 137 2.49 12.24 248 3320 68 Moongi 126 2.85 11.7 198 2830 69 Swarnakranti 124 3.42 10.86 274 3810 70 Kalojeera 99 3.69 9.96 188 2820 71 Ketekijoha 114 2.49 17.04 230 4130 72 Maudamani 109 1.85 14.55 210 7740 73 Tarori Basmati 113 2.01 15.71 218 2860 74 Mamihunger 110 2.92 20.36 175 2640 75 Sneha 112 2.31 20.36 240 3860 76 Savitri 123 2.07 12.32 275 5680 77 CR Dhan 101 94 1.86 14.28 225 4670 78 CR Dhan 907 112 3.05 15.59 280 4250 79 CR Dhan 801 114 1.92 14.84 305 6340 80 Chinikamini 116 2.46 15.04 226 3680 81 Nuakalajeera 112 2.98 17.16 210 3670 82 Moti 113 1.98 16.12 265 4760 83 Nuadhusura 114 3.05 15.82 232 3580 84 Heera 70 1.85 17.23 190 2940 Pradhan et al. BMC Plant Biology (2020) 20:57 Page 5 of 24 Table 1 Mean estimate values of days to 50% flowering, grain Fe- Zn content, panicles/m2 and grain yield of 102 genotypes including biofortified lines and check varieties studied during wet season, 2016 & 2017 (Continued) Sl. No. National testing No./Accession No./Name of the genotype Days to 50% flowering Iron content in ppm Zinc content in ppm Panicles/m2 Grain yield (kg/ha) 85 Jalmagna 130 2.32 15.76 236 3250 86 AC44756 110 2.35 15.32 180 2240 87 AC44755 112 2.23 15.63 178 2350 88 AC44754 112 2.53 15.08 192 2420 89 AC44753 110 2.13 15.73 184 2480 90 Swarna-Sub 1 116 1.81 14.48 286 6130 91 Ranjit 127 1.96 14.52 308 5620 92 Swarna 120 2.03 14.52 310 6350 93 Jaya 108 2.05 16.08 274 4580 94 Samalei 110 2.1 13.16 262 4450 95 AC44752 112 1.9 16.24 186 2520 96 Lalat 98 1.86 20.05 246 4520 97 MTU1010 97 1.79 15.04 258 4890 98 Naveen 101 1.73 14.08 268 5120 99 Satabdi 91 2.01 16.4 252 4460 100 Pooja 126 1.73 13.07 284 5470 101 Sarala 133 1.92 14.52 278 5120 102 Agnisar 105 1.75 13.46 186 2620 LSD5% 4.551 0.561 3.140 59.01 967 CV% 2.2 9.3 9.7 11.5 12.0 The two check varieties for micronutrients, Chittimuthylu and Kalanamak had mean value of 22.3 mg kg− 1 and 25.91 mg kg− 1 zinc in grain, respectively. In the studied panel, 56 genotypes showed moderate level of 15–20 mg kg− 1 grain Zn, while 32 lines exhibited low level (< 15 mg kg− 1 grain Zn). Evaluation of 102 genotypes for grain yield and Fe-Zn content over 2 years revealed two biofortified lines namely IET24779 and IET25465 showing at par yield with the yield check variety IR64 (Table 1). The genotypes containing high Fe or high Zn in milled rice and at par grain yield with yield check variety may be selected as cultivar or as donor parent for the breeding programs. The desirable genotypes with high Fe content and good grain yield were IET23829, IET 23832, BPT5204, IET24316, IET24760, IET24775, IET24779, IET25441 and IET25465. The genotypes with high Zn content and high grain yield were IET23824, IET24391, IET25446, IET25461, IET25465, IET25457, IET25477, Mumihunger, Sneha, Lalat and Chittimuthyalu. Genotypes IET24779, IET25465 and Chittimuthyalu produced at par yield with standard check variety IR64 along with high grain Fe-Zn content. The frequency of genotypes having low, moderate and high grain Fe and Zn content is depicted in spider graph and histogram (Fig. 1). Relatedness among genotypes for grain yield and Fe-Zn content through genotype-by-trait biplot analysis The scatter diagram was plotted in the first two principal components axis to generate genotype-bytrait biplot graph for grain Fe, grain Zn concentration and grain yield of the 102 genotypes present in the panel (Fig. 2). The first and second principal components had 87.82 and 8.08% of the total variability with eigen value of 10.18 and 0.937, respectively (Additional file 1: Figure S1). Among the 3 traits from the principal component analysis, the grain yield contributed maximum diversity, followed by grain Fe content and grain Zn content in the panel population (Fig. 2). From the scattering pattern of genotypes in the 4 quadrants revealed the placement of the genotypes for high grain yield and Fe content in opposite direction. Higher Zn containing genotypes are in between grain yield and Fe content. However, there are some genotypes which are located nearer to origin possessing higher estimates of grain yield and grain Fe-Zn content. These genotypes have been encircled in the figure (Fig. 2). The top right (Ist quadrant) and bottom right (2nd quadrant) possessed majority of the genotypes with better in Fe and Zn content in kernel along with Pradhan et al. BMC Plant Biology (2020) 20:57 Page 6 of 24 Fig. 1 Fe and Zn content of 102 genotypes and their frequency distribution in the panel population. a Spider graph showing the Fe and Zn content of the genotypes. b Frequency of high, moderate and low Fe-Zn genotypes in the panel population higher grain yield. The 3rd (bottom left) and 4th quadrant (top left) accommodates genotypes most of which were poor in grain Fe and Zn content with low grain yield (Fig. 2). Many genotypes with moderate in Fe and Zn content are found in 2nd quadrant. The desirable genotypes containing moderate to high grain Fe- Zn and grain yield are located both side of the X-axis and are encircled in the scatter diagram (Fig. 2). Genetic diversity in the panel population using 100 molecular markers The panel containing 102 germplasm lines exhibiting wide genetic variation for grain Fe and Zn content were genotyped using 100 molecular markers including 25 gene specific and 75 SSR markers. The loci used for genetic diversity and the calculated parameters are depicted in the table (Table 2). Two hundred forty four amplicons were obtained in toto with 2.44 average alleles per locus. Fig. 2 Genotype-by-trait biplot graph showing 102 genotypes in two main principal components for three traits. Fe: grain iron content; Zn: grain zinc content; yld: grain yield (kg/ha); PN: panicle number; DFF: days to 50% flowering. The dot numbers in the figure represent the serial number of the genotypes enlisted in Table 1 Pradhan et al. BMC Plant Biology (2020) 20:57 Page 7 of 24 Table 2 Details of 100 SSR and direct marker loci used for genotyping a panel containing 102 rice genotypes and their genetic diversity parameters Sl. No. Marker Name No. of alleles Range of amplicon (bp) Major allele frequency Gene diversity Hetero- zygosity PIC value inbreeding coefficient (f) 1 RM243 3.0000 100–140 0.6238 0.5331 0.1980 0.4705 0.6315 2 RM488 2.0000 190–210 0.7344 0.3901 0.0104 0.3140 0.9736 3 RM490 2.0000 95–115 0.6450 0.4580 0.0500 0.3531 0.8919 4 RM574 2.0000 150–160 0.8618 0.2381 0.0132 0.2098 0.9455 5 RM122 2.0000 240–260 0.6480 0.4562 0.1939 0.3521 0.5785 6 RM234 3.0000 150–160 0.6050 0.5407 0.0500 0.4704 0.9084 7 RM248 3.0000 95–115 0.3670 0.6626 0.1383 0.5884 0.7933 8 RM8007 2.0000 155–175 0.5479 0.4954 0.1809 0.3727 0.6381 9 RM17 2.0000 150–170 0.5990 0.4804 0.0521 0.3650 0.8927 10 RM260 2.0000 100–130 0.7407 0.3841 0.0370 0.3103 0.9053 11 RM7 3.0000 140–400 0.4516 0.6374 0.0000 0.5625 1.0000 12 RM517 2.0000 250–270 0.7590 0.3658 0.0000 0.2989 1.0000 13 RM501 2.0000 150–165 0.7611 0.3636 0.0556 0.2975 0.8488 14 OsZIP4 2.0000 310–310 0.6569 0.4508 0.0000 0.3492 1.0000 15 RM594 2.0000 295–310 0.8110 0.3066 0.1585 0.2596 0.4876 16 RM3412 3.0000 200–260 0.5506 0.5936 0.0225 0.5261 0.9626 17 RM5638 2.0000 200–250 0.5956 0.4817 0.0735 0.3657 0.8494 18 RM6712 3.0000 105–200 0.6071 0.5541 0.3878 0.4941 0.3049 19 RM168 2.0000 100–120 0.5700 0.4902 0.0200 0.3701 0.9596 20 RM5626 2.0000 190–210 0.8118 0.3056 0.0471 0.2589 0.8477 21 RM3392 4.0000 160–170 0.3416 0.6940 0.4653 0.6320 0.3339 22 RM1278 2.0000 130–140 0.7150 0.4076 0.1500 0.3245 0.6350 23 RM471 2.0000 110–130 0.7500 0.3750 0.1327 0.3047 0.6492 24 RM521 2.0000 240–270 0.9278 0.1340 0.0111 0.1250 0.9180 25 RM6209 2.0000 80–90 0.9798 0.0396 0.0000 0.0388 1.0000 26 RM80 2.0000 135–160 0.6198 0.4713 0.2604 0.3602 0.4516 27 OsZIP8 2.0000 0–100 0.6863 0.4306 0.0000 0.3379 1.0000 28 RM152 3.0000 140–180 0.4023 0.6585 0.5632 0.5845 0.1503 29 RM440 3.0000 150–205 0.7961 0.3438 0.0132 0.3150 0.9622 30 RM432 3.0000 170–350 0.7990 0.3399 0.0882 0.3121 0.7426 31 RM434 3.0000 150–290 0.7677 0.3706 0.0808 0.3233 0.7839 32 RM 3 3.0000 115–160 0.5337 0.5716 0.0337 0.4871 0.9417 33 RM 1 3.0000 75–120 0.7464 0.4101 0.0435 0.3730 0.8954 34 RM 144 3.0000 235–270 0.5000 0.5955 0.1000 0.5137 0.8343 35 RM 201 2.0000 140–160 0.8693 0.2272 0.0568 0.2014 0.7524 36 RM 205 4.0000 100–140 0.6325 0.5547 0.0843 0.5172 0.8497 37 RM 270 2.0000 300–370 0.9412 0.1107 0.0000 0.1046 1.0000 38 RM 335 2.0000 100–110 0.6616 0.4478 0.0303 0.3475 0.9330 39 RM154 3.0000 130–200 0.4765 0.6358 0.4471 0.5641 0.3022 40 RM211 3.0000 120–170 0.6534 0.4884 0.1591 0.4156 0.6774 41 RM202 3.0000 150–200 0.5150 0.6106 0.0900 0.5385 0.8540 42 RM293 2.0000 200–210 0.7557 0.3693 0.0114 0.3011 0.9696 43 RM85 3.0000 90–110 0.5058 0.6182 0.0814 0.5468 0.8698 Pradhan et al. BMC Plant Biology (2020) 20:57 Page 8 of 24 Table 2 Details of 100 SSR and direct marker loci used for genotyping a panel containing 102 rice genotypes and their genetic diversity parameters (Continued) Sl. No. Marker Name No. of alleles Range of amplicon (bp) Major allele frequency Gene diversity Hetero- zygosity PIC value inbreeding coefficient (f) 44 RM407 6.0000 170–550 0.4892 0.7019 0.1290 0.6713 0.8180 45 RM237 2.0000 130–150 0.5345 0.4976 0.0115 0.3738 0.9772 46 RM259 4.0000 180–280 0.6867 0.4806 0.0400 0.4346 0.9178 47 RM421 2.0000 240–300 0.9521 0.0912 0.0532 0.0870 0.4209 48 RM235 3.0000 90–150 0.6462 0.4857 0.2462 0.4044 0.4990 49 RM1337 2.0000 120–140 0.6875 0.4297 0.5000 0.3374 ###### 50 RM3409 2.0000 304–340 0.7692 0.3550 0.0000 0.2920 1.0000 51 RM105 2.0000 130–145 0.6076 0.4768 0.0000 0.3632 1.0000 52 RM309 1.0000 190–190 1.0000 0.0000 0.0000 0.0000 NaN 53 RM452 1.0000 250–250 1.0000 0.0000 0.0000 0.0000 NaN 54 RM204 3.0000 120–170 0.4695 0.5741 0.0244 0.4807 0.9580 55 RM137 1.0000 250–250 1.0000 0.0000 0.0000 0.0000 NaN 56 RM1789 1.0000 180–180 1.0000 0.0000 0.0000 0.0000 NaN 57 RM6641 2.0000 140–150 0.7527 0.3723 0.0000 0.3030 1.0000 58 RM296 4.0000 140–410 0.5337 0.6233 0.0337 0.5673 0.9465 59 RM3331 2.0000 165–175 0.5306 0.4981 0.0000 0.3741 1.0000 60 RM31 3.0000 100–125 0.5000 0.6157 0.0455 0.5419 0.9270 61 RM429 2.0000 120–140 0.9511 0.0930 0.0326 0.0887 0.6527 62 RM556 2.0000 130–190 0.7157 0.4070 0.0196 0.3242 0.9523 63 RM585 4.0000 130–210 0.3824 0.6915 0.0196 0.6330 0.9719 64 RM23 2.0000 150–160 0.8713 0.2243 0.0000 0.1991 1.0000 65 RM34 2.0000 160–180 0.9188 0.1493 0.0875 0.1382 0.4191 66 RM53 2.0000 180–200 0.5714 0.4898 0.0000 0.3698 1.0000 67 RM300 3.0000 130–150 0.6344 0.5305 0.0000 0.4743 1.0000 68 RM315 2.0000 145–150 0.8687 0.2281 0.0000 0.2021 1.0000 69 RM339 3.0000 150–190 0.7475 0.4081 0.0707 0.3706 0.8283 70 RM400 4.0000 250–340 0.6571 0.5224 0.0000 0.4819 1.0000 71 RM528 2.0000 300–320 0.5464 0.4957 0.0000 0.3728 1.0000 72 RM486 2.0000 135–140 0.7526 0.3724 0.0206 0.3031 0.9452 73 RM340 6.0000 130–290 0.5215 0.6656 0.1828 0.6293 0.7279 74 RM1132 4.0000 95–140 0.5505 0.6145 0.0303 0.5617 0.9512 75 RM441 1.0000 170–170 1.0000 0.0000 0.0000 0.0000 NaN 76 RM590 1.0000 190–190 1.0000 0.0000 0.0000 0.0000 0.0000 77 RM258 1.0000 250–250 1.0000 0.0000 0.0000 0.0000 0.0000 78 GRMM9–1 2.0000 0–275 0.6961 0.4231 0.0000 0.3336 1.0000 79 GRMM9–2 2.0000 0–150 0.9216 0.1446 0.0000 0.1341 1.0000 80 OsNAC 2.0000 0–600 0.7917 0.3299 0.0000 0.2755 1.0000 81 OsZIP8A 2.0000 0–900 0.5882 0.4844 0.0000 0.3671 1.0000 82 OsZIP8C 2.0000 0–930 0.5980 0.4808 0.0000 0.3652 1.0000 83 OsYSL4E 3.0000 0–851 0.5000 0.6250 0.0000 0.5546 1.0000 84 OsMTP1A 2.0000 0–950 0.7451 0.3799 0.0000 0.3077 1.0000 85 OsNRAMP5G 2.0000 0–150 0.6373 0.4623 0.0000 0.3555 1.0000 86 IRMM9–1 2.0000 0–200 0.7059 0.4152 0.0000 0.3290 1.0000 Pradhan et al. BMC Plant Biology (2020) 20:57 Page 9 of 24 Table 2 Details of 100 SSR and direct marker loci used for genotyping a panel containing 102 rice genotypes and their genetic diversity parameters (Continued) Sl. No. Marker Name No. of alleles Range of amplicon (bp) Major allele frequency Gene diversity Hetero- zygosity PIC value inbreeding coefficient (f) 87 OsYSL1 2.0000 0–230 0.6667 0.4444 0.0000 0.3457 1.0000 88 OsYSL2A 2.0000 0–150 0.9216 0.1446 0.0000 0.1341 1.0000 89 OsYSL2B 3.0000 0–330 0.4412 0.6353 0.0000 0.5590 1.0000 90 OsYSL5 2.0000 0–380 0.5098 0.4998 0.0000 0.3749 1.0000 91 OsYSL6 2.0000 0–166 0.5098 0.4998 0.0000 0.3749 1.0000 92 OsYSL11 2.0000 0–190 0.6569 0.4508 0.0000 0.3492 1.0000 93 OsZIP6A 3.0000 0–170 0.5098 0.5352 0.0000 0.4280 1.0000 94 OsZIP6B 2.0000 0–220 0.6078 0.4767 0.0000 0.3631 1.0000 95 OsZIP7 3.0000 0–180 0.6569 0.4796 0.0000 0.4025 1.0000 96 OsZIP8 2.0000 0–160 0.5294 0.4983 0.0000 0.3741 1.0000 97 OsNRAMP1A 3.0000 0–310 0.6667 0.4931 0.0000 0.4364 1.0000 98 OsNRAMP1B 2.0000 0–250 0.6471 0.4567 0.0000 0.3524 1.0000 99 OsFER1 3.0000 0–170 0.8431 0.2737 0.0000 0.2518 1.0000 100 OsFER2 3.0000 0–475 0.4804 0.5992 0.0000 0.5157 1.0000 Mean 2.44 0.682038 0.414195 0.061385 0.348211 0.849347 The number of alleles ranged from 1 to 6 per marker with RM340 showing the highest number of alleles in the panel genotypes. The average major allele frequency of Fe and Zn linked polymorphic markers was observed to be 0.675 ranging within the bracket of 0.4023 (RM152) and 1.0000 (RM309, RM452, RM137, RM1789 and RM441) (Table 2). The PIC value ranged from 0.000 (RM6209, sZIP8, RM3409, RM309, RM452, RM137, RM1789, RM556, RM23, RM34, RM441 and GRMM9– 2) to 0.6713 (RM407) with mean PIC of 0.3553. The observed average heterozygosity (Ho) was 0.0496 with a range of 0.00–0.6714. Amongst the markers used, only 36 showed heterogygocity value to be more than zero, whereas 64 exhibited zero values. The average gene diversity (He) varied from 0.0000 (RM309, RM452, RM137, RM1789, RM44, RM590 and RM258) to 0.6711 (RM3392) with an average value of 0.35. Population structure A Bayesian clustering approach was followed for estimation of genetic structure by taking probable subpopulations (K) and higher delta K-value using the STRUCTURE 2.3.6 software. The genotypes in the panel exhibiting variation for grain Fe and Zn content were evaluated for genetic structure following. It categorized the genotypes into two sub-populations (Fig. 3a, b) with a high ΔK peak value of 495.2 at K = 2 among the assumed K (Fig. 3a). But, the classification was not robust with respect to Fe-Zn content in the genotypes. However, at K = 3 with ΔK value of 117.9 categorized the population into three distinct sub-populations with perfect grouping. At K = 3, the sub-population 1 (SP1), sub-population 2 (SP2) and sub-population 3 (SP3) included 18, 44 and 40 genotypes, respectively whereas eleven genotypes were admixture type having major genetic constituent from three sub-groups (Fig. 3c; Table 3). The SP1 and SP2 included majority of moderate and high grain Fe-Zn containing genotypes accommodating 100% of total high Fe (> 4 mg kg− 1) genotypes, 97.3% of total moderate Fe (3–4 mg kg− 1) genotypes, 90% of total high Zn (> 20 mg kg− 1) genotypes and 69.2% of total moderate Zn (15–20 mg kg− 1) genotypes. Maximum allele frequency divergence between the two subpopulations (net nucleotide distance) 1–2, 1–3 and 2–3 were 0.1239, 0.1813 and 0.2480, respectively. The average distance (expected heterozygosity) between individuals in cluster 1, cluster 2 and cluster 3 were 0.3140, 0.3051 and 0.3115, respectively. The three sub-populations showed fixation index values (FST) of 0.287 for SP1, 0.385 for SP2 and 0.4853 for SP3. A lower value of alpha (alpha = 0.0388) was found for the studied panel population. The distribution pattern of alpha-value in the panel population showed a leptokurtic symmetry while distribution of FST values in the sub-populations were almost in symmetric shape with same to the left and right from the centre (Additional file 2: Figure S2). Further at K = 7 with ΔK value of 3.946, the software categorized the entire population into 7 distinct sub-populations classifying the population into more subtle classes according to grain Fe-Zn content. Pradhan et al. BMC Plant Biology (2020) 20:57 Page 10 of 24 Fig. 3 a Graph of delta K value, an ad-hoc statistic related to the rate of change in the log probability of data between successive K values; b Population structure of the 102-panel population placed based on membership probability fractions of individual genotypes at K = 2 and c Population structure of the 102-panel population placed based on membership probability fractions of individual genotypes at K = 3. The genotypes with the probability of ≥80% membership fractions were assigned to corresponding subgroups with others categorized as admixture. The numbers in the figure represent the serial number of the genotypes enlisted in Table 1 Analysis of molecular variance (AMOVA) and LD decay plot The analysis of molecular variance (AMOVA) showed genetic variations between and within the subpopulations at K = 2, K = 3 and K = 7 (Table 4). The genetic variations between and within the two subpopulations (K = 2) was 41% among the populations, 50% among individuals and 9% variation within individuals in the panel population. At 3 subpopulations, 38% of the variation among populations, 52% among individuals and 10% variation within individuals in the panel population. But, the analysis at K = 7 revealed 39% of the variation among populations, 51% among individuals and 10% variation within individuals resulted from the analysis. Wright’s F statistic was used to calculate the deviation from Hardy-Weinberg’s