Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/oxford-university-press/high-throughput-genotype-based-population-structure-analysis-of-SNuHAut90u
- Copyright
- Copyright © 2021 American Society of Animal Science
- eISSN
- 2573-2102
- DOI
- 10.1093/tas/txab033
- Publisher site
- See Article on Publisher Site
Abstract
Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 High-throughput genotype-based population structure analysis of selected buffalo breeds † † † ‡ || || Prakash B. Thakor, Ankit T. Hinsu, Dhruv R. Bhatia, Tejas M. Shah, Nilesh Nayee, A. Sudhakar, † ‡,$,1 Dharamshibhai N. Rank, and Chaitanya G. Joshi Department of Animal Genetics and Breeding, College of Veterinary Science and Animal Husbandry, Anand Agriculture University, Anand 388001, India; Department of Animal Biotechnology, College of Veterinary || Science and Animal Husbandry, Anand Agriculture University, Anand 388001, India; National Dairy Development Board, Anand 388001, India; and Gujarat Biotechnology Research Centre, Gandhinagar 382017, India ABSTRACT: India is considered as the home of Surti, Pandharpuri, and Jaffarabadi in first tract of some of the best buffalo breeds. two principal components and at K = 4, respect- However, the genetic structure of the Indian ively, while remaining breeds were grouped to- river buffalo is poorly understood. Hence, there gether as a separate single cluster and admixed. is a need to characterize the populations and Murrah and Mehsana showed early linkage understand the genetic structure of various buf- disequilibrium (LD) decay, while Surti breed falo breeds for selection and to design breeding showed late decay. In LD blocks to quantitative strategies. In this study, we have analyzed trait locis (QTLs) concordance analysis, 4.65% genetic variability and population structure of concordance was observed with 873 LD of seven buffalo breeds from their respective blocks overlapped with 2,330 QTLs. Overall, geographical regions using Axiom Buffalo total 4,090 markers were identified from all Genotyping Array. Diversity, as measured by LD blocks for six types of traits. Results of expected heterozygosity, ranged from 0.364 in this study indicated that these single-nucleotide Surti to 0.384 in Murrah breed, and pair-wise polymorphism (SNP) markers could differen- F values revealed the lowest genetic distance tiate phenotypically distinct breeds like Surti, ST between Murrah and Nili-Ravi (0.0022), while Pandharpuri, and Jaffarabadi but not others. the highest between Surti and Pandharpuri So, there is a need to develop SNP chip based (0.030). Principal component analysis and on SNP markers identified by sequence infor- structure analysis unveiled the differentiation mation of local breeds. Key words: admixture, affymetrix, Indian buffalo, linkage disequilibrium, quantitative trait loci, single-nucleotide polymorphism © The Author(s) 2021. Published by Oxford University Press on behalf of the American Society of Animal Science. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribu- tion, and reproduction in any medium, provided the original work is properly cited. Transl. Anim. Sci. 2021.5:1-14 doi: 10.1093/tas/txab033 Corresponding author: cgjoshi@rediffmail.com Received June 15, 2020. Accepted May 7, 2021. 1 Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. INTRODUCTION by neutral molecular markers, such as SNPs and microsatellites, allows the determination of their The importance of genetic diversity in livestock extinction risk and to design strategies for their is directly related to the need for genetic improve- management, conservation, and maintenance of ment of economically important traits as well as genetic variation for continuous genetic improve- to facilitate rapid adaptation to potential changes ment (Toro and Maki-Tanila, 2007). The river buf- as per breeding goals. Population structure, and falo has been selected as a dairy animal with several unusual levels of shared ancestry, can potentially recognized breeds, spread from the Indian sub- cause problems with genome-wide association stud- continent to the eastern Mediterranean countries ies. The analysis of a large number of SNPs across (the Balkans, Italy, and Egypt). For overall breed the genome will reveal aspects of the population improvement and to meet future challenges, imme- genetic structure, including evidence of adaptive diate action is required for the characterization of selection across the genome (Barendse et al., 2009). buffalo breeds in India. Comprehensive knowledge Domestication of animals have changed the mor- of genetic variation within and among different phological and behavioural characteristics through breeds is necessary for understanding and improv- selection programmes for improving the produc- ing traits of economic importance. Current study tion traits. That ultimately leads to the formation was performed based on SNP genotyping data to of very diverse breeds (Diamond, 2002; Toro and determine the genetic structure of Indian buffalo Maki-Tanila, 2007; Flori et al., 2009). breeds so as to construct appropriate conservation River buffalo domestication is likely to have strategies and to utilize the breed variation. occurred around 6,300 years before present in north-western India (Kumar et al., 2007; Nagarajan MATERIALS AND METHODS et al., 2015). India is one of the largest milk pro- ducer in the world, producing over 155.5 million Animals and Sampling ton milk during 2015–2016 and about 49% of milk production is contributed by buffaloes (FAO, 2007; A total of 295 female buffaloes from seven Department of Animal Husbandry, Fisheries and breeds were used in this study (Table 1). All animals Dairying, Government of India, 2015). India has were selected based on their true breed-specific mor - approximately 108.7 million buffaloes (Department phological traits from their respective home tract of Animal Husbandry, Fisheries and Dairying, (Nayee et al., 2016), avoiding sampling from related Government of India, 2015) with 13 registered animals (Supplementary Fig. S1). Blood samples breeds recognized based on their phenotypic were collected from all the selected animals. This traits, production performance, utility pattern, and work was ethically approved by the Institutional eco-geographical distribution. Animal Ethic Committee (IAEC) of College of Genetic analysis is facilitated by genotyping Veterinary Sciences and A.H., Anand Agricultural polymorphic genetic loci, also called as genetic University, Anand (letter no.: IAEC 155/2011). variants or markers. SNPs are the most common type of genetic variants, consisting of single-nu- SNP Genotyping cleotide differences between two individuals at a particular site in the DNA sequence. Assessing DNA was extracted using Qiagen QIAamp genetic biodiversity and population structure of Blood DNA kit (Qiagen, Germany) as per manu- minor breeds through the information provided facturer’s instructions. DNA quantity and quality Table 1. Summary of genotyped samples Sr. no. Name of breed Sample origin/breeding tract Sample size 1 Murrah Haryana: Rohtak, Hissar, and Jind 70 Punjab: Nabha and Patiala 2 Nili-Ravi Punjab: Amritsar, Gurdaspur, and Ferozepur 33 3 Mehsana North Gujarat: Mehsana, Banaskantha, and Patan 75 4 Jaffarabadi Saurashtra, Gujarat: Amreli, Gir, Junagadh, Bhavnagar, and Rajkot 41 5 Banni Gujarat: Kutchh 20 6 Pandharpuri Southern Maharashtra: Solapur, Satara, and Latur 34 7 Surti Southern Gujarat: Anand, Kheda, Baroda, and Surat 22 Total 295 Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo were checked using Nanodrop (Thermo Fisher clustering based on genetic distance (Fitch and Scientific, MA) and agarose gel electrophoresis, re- Margoliash, 1967). spectively. SNP genotyping was carried out using Several statistical parameters were stated to commercially available Axiom Buffalo Genotyping measure the extent of LD. The r is a better de- Array (90 K) designed with 123,040 probes on scriptor of LD as it is more robust and not sensitive Gene Titan MC (Thermo Fisher Scientific, MA) in- to changing gene frequency and effective popula- strument at a commercial laboratory (Imperial Life tion size (Zhao et al., 2007). Effective population Science Group, Gurgaon). It was designed based size can be estimated for several past generations on SNPs discovered from Mediterranean, Murrah, for given population using the available informa- Jaffarabadi, and Nili-Ravi breeds of buffaloes. In tion of correlation between gene frequencies and the designed array, there are 123,040 probes, which LD (Sved, 1971). The decline rate of LD with include 89,988 probes for SNPs, while other probes intermarker distance was estimated using bin size are for quality control (QC) and gender calling. of 10 kb distance between SNPs. Breed-wise effective population size (Ne) was cal- culated using SNeP v1.1 (Barbato et al., 2015) with Data Filtering and Quality Control parameters: bin-width = 50,000 bp; minimum distance Only SNPs mapped to autosomal chromo- between SNPs = 50,000 bp, maximum distance be- somes were used in this study. Data were filtered with tween SNPs = 4,000,000 bp and minimum allele fre- PLINK v1.07 (Purcell et al., 2007) based on criteria: quency = 0.05. SNeP estimates Ne from genome-wide removal of SNPs with same UMD position, missing LD using the method suggested by Corbin et al. (2012). genotypes (<0.1), minor allele frequency (<0.05) and Population clustering was performed using principal Hardy–Weinberg Equilibrium (P-value <0.00001; component analysis in order to place the breed groups Supplementary Table S1). None of the animal/in- with respect to their genetic constitutes with PLINK-1.9 dividual was removed during the quality filtering, (Chang et al., 2015) using 285 highly variable markers whereas 75,704 SNPs remained after filtering. (allele frequency difference between breeds >0.5) and plotted using scatterplot3d package in R. Breed struc- ture and breed differentiation was performed using Genetic Diversity Assessment fastSTRUCTURE (Raj et al., 2014) using same 285 Observed and expected genotype frequencies highly variable markers. The differentiation of popu- within each breed was calculated for all the loci lations was performed up to the group (K) level of 8 using PLINK v1.07 and the results were evalu- using simple model. The fastSTRUCTURE analysis ated based on P-values for significance test for provided ancestry proportions for each sample under Hardy-Weinberg Equilibrium, obtained for each analysis, which was graphically represented by distruct. loci. Linkage disequilibrium (LD) was calculated py script within the fastSTRUCTURE software. using PLINK and r values were calculated for all SNP pairs that were located less than 1,000 SNPs Genome-Wide LD Block Mapping on Quantitative apart and falling under 10 Mb distance windows. Trait Locis (QTLs) Furthermore, SNPs were binned with bin size of Linkage disequilibrium blocks, combination of 10,000 bases distance, and average r value of each alleles linked along a chromosome and inherited to- bin was plotted against median distance value gether from a common ancestor, were generated ggplot2 v2.2.1 package in R v3.3. Pair-wise F ST with Java-based gPLINK v1.0 and Haploview v2.01 values and the associated 95% confidence inter - (Barrett et al., 2004). Blocks were defined by employing vals were calculated using the Hierfstat package haplotypic diversity criterion, where a small number (Goudet, 2005) in R. Wright’s inbreeding coefficient of common haplotypes provide high chromosomal estimated as F , which is caused by Wahlund ef- IS frequency coverage (Patil et al., 2001; Zhang et al., fect by mixing individuals from genetically different 2002, 2003; Anderson and Novembre, 2003). The al- populations, and normalized variance in allele fre- gorithm suggested by Gabriel et al. (2002) was used, quencies between populations is estimated as F ST which defines a pair of SNPs to be in strong LD if (Zhivotovsky, 2015). Pair-wise F values between ST the upper 95% confidence bound of D′ value is be- all possible combinations of breeds were estimated tween 0.7 and 0.98. Reconstructed haplotypes were and subsequently phylogenetic tree was generated inserted into Haploview v2.01 to estimate LD stat- in Fitch–Phylip using Fitch–Margoliash method, istics and construct the blocking pattern for all 29 which uses a weighted least squares method for Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. autosomes. LD blocks were estimated using an accel- QC measures, 295 samples and 75,704 SNPs re- erated Expectation–Maximization algorithm method mained for population analysis (Supplementary described by Qin et al. (2002). QTL database was re- Table S1). trieved from previously reported QTLs in Animal QTLdb (Hu et al., 2013). QTL data set of cattle (Bos Allele Frequency-Based Differentiation taurus) QTL_UMD_3.1.1 was used as a reference for Highest number of SNPs with alternate allele the analysis, containing the information regarding six frequency between 0.3 and 0.4 was observed in types of the traits: milk traits; health traits; production all studied buffalo breeds except Surti (Fig. 1A). traits; reproduction traits; exterior traits; and meat and Highest allele count was observed in the range of carcass traits. The QTL files were intersected with the frequency class 0.2–0.5. Highest average alternate les of fi LD blocks using Bedtools v2.26.0 (Quinlan and allele frequency was observed in Nili-Ravi (0.3051), Hall, 2010) to obtain information of QTLs overlapping while Jaffarabadi showed least (0.3028) among all with LD blocks. breeds (Supplementary Fig. S2). The distribution of alternate allele did not significantly differ be- RESULTS tween studied breeds. Highest proportion of alter- nate alleles was observed in Murrah with 91.86%, Genetic Diversity Analysis while lowest proportion was observed in Surti with Samples were genotyped with the average call 89.86% (Fig. 1B). The observed heterozygosity (H ) rate of passed sample 98.58%. Upon applying and expected heterozygosity (H ) in all breeds did Figure 1. Alternate allele distribution. (A) Distribution of alternate allele frequency in studied buffalo breed. (B) Breed-wise proportion and distribution of alternate allele with allele frequency >0 (monomorphic SNPs were removed; BBN: Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: Mehsana, BPN: Pandharpuri, BST: Surti). Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo not differ and ranged from 0.3719 (Pandharpuri) the number of groups). For K = 3, total three breed to 0.3864 (Murrah) and 0.3643 (Surti) to 0.3846 groups were identified ( Fig. 4); cluster I included (Murrah), respectively (Table 2). The lowest F Banni, Murrah, Mehsana, and Nili-Ravi breeds; IS was observed for Murrah (−0.0046) followed by cluster II included Jaffarabadi breed only; and Mehsana (−0.0070), while comparative higher cluster III grouped Pandharpuri and Surti breeds. values were observed in Surti (−0.0314) followed by But when K = 4 was assumed, cluster III further Banni (−0.0270). split into independent cluster for Surti breed. These clusters obtained were consistent with the neigh- bor-joining tree. The membership of Cluster I was F -Based Differentiation ST consistent with breed histories, with one cluster F values showed least genetic distance be- including a pair of closely related breeds (Murrah ST tween Murrah and Nili-Ravi (0.00221) followed and Mehsana), which showed some level of admix- by Murrah and Mehsana (0.00402), while highest ture. It seemed to be an optimum of four clusters, genetic distance was observed between Surti and which was also indicated by a maximum likeli- Pandharpuri (0.03097) followed by Surti and Banni hood method (Supplementary Fig. S4). So, K = 4 (0.02650; Supplementary Table S2). Based on F was considered to represent most relevant number ST values, neighbor-joining tree placed Nili-Ravi and of genetic clusters in the data sets, which corres- Murrah, as well as Mehsana and Banni together ponded to their breed designation. Surti breed in two separate clusters, which corresponds with showed better separation with small amount of ad- their geographical origin (Fig. 2). Furthermore, mixture at all levels, while Murrah and Mehsana this clustering pattern was also supported by neigh- bor-joining tree generated using studied SNPs (Supplementary Fig. S3). This differentiation also correlates with the morphological differentiation of the buffalo breeds. Principal Component Analysis (PCA) Results The total variability explained by first three principal components was 65.6%, of which first, second, and third components explained 30.05%, 27.14%, and 8.45%, respectively. This variation re- sulted in a separate cluster of Surti, Pandharpuri, and Jaffarabadi on coordinates 1, 2, and 3, respect- ively, while other breeds remain admixed (Fig. 3). Model-Based Population Assignment Furthermore, relatedness between breeds and Figure 2. Phylogenetic tree of breed differentiation based on pair- the significance of the existence of subpopulations wise F values. Labeled tree with the name of breed at each leaf (BBN: ST was investigated by model-based unsupervised Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: clustering using K = 2 to K = 8 (K values indicates Mehsana, BPN: Pandharpuri, BST: Surti). Table 2. Genetic diversity parameters in Indian buffalo breeds from genotyped data (BBN: Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: Mehsana, BPN: Pandharpuri, BST: Surti) Number of Observed heterozy- Expected heterozy- Inbreeding coeffi- Breed Animals gosity, H (mean ± SE) gosity, H (mean ± SE) cient, F (mean ± SE) O E IS BBN 20 0.3839 ± 0.0006 0.3738 ± 0.0005 −0.0270 ± 0.0036 BMS 75 0.3857 ± 0.0005 0.3830 ± 0.0005 −0.0070 ± 0.0033 BNR 33 0.3832 ± 0.0006 0.3799 ± 0.0005 −0.0089 ± 0.0072 BPN 34 0.3719 ± 0.0006 0.3680 ± 0.0005 −0.0107 ± 0.0116 BJF 41 0.3839 ± 0.0006 0.3738 ± 0.0005 −0.0098 ± 0.0031 BMR 70 0.3864 ± 0.0005 0.3846 ± 0.0005 −0.0046 ± 0.0024 BST 22 0.3757 ± 0.0007 0.3643 ± 0.0005 −0.0314 ± 0.0094 Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. Figure 3. 2D principal component analysis (PCA) plot of all seven buffalo breeds together up to principal components 5 (BBN: Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: Mehsana, BPN: Pandharpuri, BST: Surti). breed showed higher amount of admixture con- In Surti breed, LD decayed late as distance between sistent with its crossing with other breeds. With loci increased compared to other breeds. However, increasing K values, Pandharpuri and Surti showed Mehsana and Murrah showed early decay among separation at all subsequent levels (Fig. 4). Three all the breeds. Jaffarabadi individuals were identified as pure A continuous steady decline in effective popu- breed based on Q-value greater than 95%, while lation size was observed over the last 1,000 gen- the remaining showed variable amount of admix- erations in all breeds. Effective population size ture. Similarly, Pandharpuri buffaloes showed the of Murrah and Mehsana has drastically declined highest number (26) of purebred individuals with over the last 100 generations with steeper slope, more than 80% of Q-value. Likewise, Surti breed while Surti and Banni are declining at lower rate has 19 purebred individuals with negligible admix- (Fig. 5B). Jaffarabadi, Nili-Ravi, and Pandharpuri ture with other breeds. showed intermediate rate of declination over the last 100 generations. LD Analysis Genome-Wide Study of LD blocks LD decay showed highest r value in Surti (from 0.412 to 0.175) followed by Banni (from 0.412 to Total 1,144 LD blocks were obtained with 0.169; Fig. 5A), while Pandharpuri (from 0.379 to the highest number of blocks on chromosome 1 0.149) and Nili-Ravi (from 0.412 to 0.139), as well as (99 blocks), while the least number of blocks on Mehsana (from 0.378 to 0.128) and Murrah (from chromosome 28 with 19 blocks (Table 3). Overall, 0.382 to 0.120), decayed almost with the same rate. the mean number of SNPs in block ranged from Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo Figure 4. Estimated population structure by fastSTRUCTURE for K = 2 to K = 8. Each individual is represented by a thin vertical line, and each breed is demarcated by a thick vertical black line (BBN: Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: Mehsana, BPN: Pandharpuri, BST: Surti). 2.75 to 4.54 SNPs per chromosome, while the max- Furthermore, analysis was also performed imum number of SNPs per block ranged from 5 based on markers overlapping with QTLs of milk (chromosome 18) to 16 (chromosome 17). Overall, fat percentage (143 markers) and body weight (315 frequency-based size distribution of LD blocks re- markers) using phenotypic recorded data from vealed that the highest number (547) of LD blocks National Dairy Development Board (India) (www. were found having sizes less than 50 kb, while very nddb.coop) and Central Institute for Research few (15) were observed having sizes as high as 450– on Buffalo (India), (cirb.res.in) respectively. 499 kb (Fig. 6). Surprisingly, no particular pattern was observed LD blocks—QTL concordance. Out of 1,144 linking phenotypic data (literature-based QTLs of LD blocks (4,090 markers), 436 (1,624 markers), milk fat and body weight) with trait-specific mark- 368 (1,285 markers), 326 (1,253 markers), 345 (1,351 er-based separation. markers), 104 (426 markers), and 81 LD blocks (338 markers) overlapped with QTLs for traits, DISCUSSION such as milk production (Supplementary Fig. S5), production (Supplementary Fig. S6), reproduc- Genetic diversity studies conducted for buffalo in tion (Supplementary Fig. S7), meat and carcass India have previously relied primarily on the use of (Supplementary Fig. S8), health (Supplementary microsatellites markers (Pundir et al., 2000; Kumar Fig. S9), and exterior (Supplementary Fig. S10), re- et al., 2006; Tantia et al., 2006; Kataria et al., 2009; Joshi spectively. Concordance, measured as a proportion et al., 2013; Joshi et al., 2015), while the use of SNP of LD blocks and QTLs overlapping each other, was genotype data in Indian cattle has also been previously highest on chromosome 1 (16.91%) and lowest on reported (Dash et al., 2017). Previously, Perez-Pardal chromosome 14 (0.91%). Overall, the concordance et al. (2018) have performed the study on 15 buffalo of all the chromosomes together was 4.65%, with animals each from river and swamp buffalo using cattle 873 LD blocks intersected with 2,330 QTLs (Table SNP array (Illumina BovineHD BeadChip) and they 4). Chromosome-wise distribution of LD blocks, have confirmed that analysis has better suitability for number of markers, and mapped QTLs for respective population structure, hybridization, and breed identi- traits is shown in Supplementary Table S3. fication of water buffalo populations. Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. Figure 5. LD study of Buffalo breeds: (A) LD decay plot based on all pair-wise comparisons between adjacent loci of all seven breeds. The horizontal axis depicts the intermarker distance in base pair and vertical axis shows the average r values. (B) Effective population size (Ne) of dif- ferent breeds with respect to generation time (BBN: Banni, BJF: Jaffarabadi, BMR: Murrah, BNR: Nili-Ravi, BMS: Mehsana, BPN: Pandharpuri, BST: Surti). The chip used in this study was designed based breeds show the genetic diversity that exists within on SNP markers of four breeds (Mediterranean, and between the breeds. Murrah and Mehsana had Murrah, Nili-Ravi, and Jaffarabadi) although using the highest number of SNPs with intermediate the reference of Bos taurus (UMD_3.1 assembly) class of frequency, suggesting that this array could (Iamartino et al., 2013). The differences in allele be utilized for these breeds for association stud- frequencies among the breeds may be caused by ies. The higher genetic variability observed in the genetic drift, adaptation to selection, or ancient di- Murrah and Mehsana, which is evident from the vergence among founder populations (MacEachern population structure analysis, suggests the intro- et al., 2009; Dadi et al., 2012). Therefore, these gression of these breeds with other breeds, such SNPs identified in this study will be useful for the as Banni, Nili-Ravi, and Jaffarabadi, while Surti study on breed structure identification and popu- and Pandharpuri showed less polymorphic SNPs, lation differentiation. Here, we used the term “al- suggesting less genetic variability. These findings ternate allele” in place of “minor allele” because are further supported by H and H values, which O E minor allele frequency does not exceed over 0.5, were found to be higher in Murrah and Mehsana while, in this study, the allele frequency exceeds breeds as compared to other breeds, which could over 0.5, often called as “fixed allele,” and, hence, be due to the availability of large population of it has been considered as an “alternate allele.” The these breeds owing to their higher milk produc- differences in observed allele frequencies among tion potential, whereas other breeds are limited in Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo numbers. Pandharpuri and Surti showed less gen- previously reported in other studies of cattle and etic variability with the lowest H suggesting that buffalo using microsatellites (Machado et al., 2003; inbreeding in conjunction with a small population Sraphet et al., 2008; Suh et al., 2014) and using size and resulted in a loss of variation within the SNP panels (Dash et al., 2017). breed. This type of low diversity among breeds was In this study, the mean F indicated that a pair ST of Surti and Pandharpuri population has greater genetic distance than other pairs. Pair-wise F ST Table 3. Chromosome-wise LD block distribution between these buffalo breeds was significantly dif- statistics with total number of LD blocks, average ferent from zero (P < 0.05). Genetic differentiation block size, mean, and maximum number of SNPs (pair-wise F ) indices observed in the present study ST in blocks are sufficient to explain the fact that these buffalo breeds are geographically well separated from each Chromo- Total LD Mean number of Max. number of other, and we had reported a similar observation some blocks SNPs per block SNPs in blocks in our previous study among Western-Central 1 99 3.48 7 Indian cattle breeds (Shah et al., 2013). Our F - 2 87 3.68 9 ST 3 59 3.25 6 based genetic classification was in agreement with 4 58 3.44 8 this classification of buffaloes except the separation 5 63 3.73 15 of Jaffarabadi breed. However, the results failed 6 43 3.72 9 to explain the hypothesis that Mehsana breed has 7 44 3.72 15 been developed using Murrah bulls on local Surti 8 52 3.75 10 buffaloes (Pundir et al., 2000) as both the breeds 9 39 4.00 8 clustered separately. Earlier study based on micro- 10 36 3.94 6 satellite markers revealed genetic diversity (F ) ST 11 54 3.51 9 based clustering between Mehsana with Jaffarabadi 12 37 3.75 9 and Surti with Pandharpuri (Kumar et al., 2007). 13 38 3.34 9 Our study also showed clustering among Surti 14 31 2.93 13 and Pandharpuri, while Mehsana and Jaffarabadi 15 33 3.00 6 16 44 3.56 12 formed separate clusters. 17 30 3.83 16 The results of the PCA analysis revealed 18 24 3.04 5 the higher amount of genetic similarities among 19 31 4.54 11 Murrah, Mehsana, Banni, and Nili-Ravi, while 20 23 3.47 9 Surti, Jaffarabadi, and Pandharpuri showed 21 36 3.94 11 greater genetic differentiations with three distinct 22 26 3.76 13 clusters. The clustering of populations from both 23 22 3.72 7 the PCA and fastSTRUCTURE indicated low 24 27 2.96 7 levels of within-population diversity of the Surti, 25 29 2.75 9 Jaffarabadi, and Pandharpuri breeds and higher di- 26 16 3.56 8 vergences of these populations from the Murrah, 27 23 3.34 8 Mehsana, Banni, and Nili-Ravi breeds. In the cur- 28 19 3.84 7 29 22 3.77 10 rent study, Surti, Jaffarabadi, and Pandharpuri All 1,145 3.56 grouped in separate clusters, contrary to earlier Figure 6. LD blocks distribution based on the size of block in respective class of size (in kb). Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. Table 4. Chromosome-wise distribution of LD blocks and QTLs with its percentage of concordance and discordance No. of No. of QTLs over- No. of LD No. of LD blocks Concordance between Chromosome QTLs lapped by LD blocks blocks overlapped with QTLs QTL and LD blocks in % 1 2,403 325 99 98 16.91 2 2,711 163 87 56 7.83 3 2,780 55 58 43 3.45 4 4,440 31 58 21 1.16 5 3,534 103 63 56 4.42 6 10,483 237 43 41 2.64 7 2,089 63 44 41 4.88 8 1,177 55 52 45 8.14 9 1,289 61 39 21 6.17 10 1,839 78 36 26 5.55 11 3,163 118 54 34 4.72 12 1,046 60 37 26 7.94 13 1,775 101 38 25 6.95 14 7,293 38 31 29 0.91 15 1,050 32 33 32 5.91 16 1,236 63 44 37 7.81 17 1,548 47 30 26 4.63 18 1,233 27 24 21 3.82 19 1,735 73 31 18 5.15 20 2,914 140 23 21 5.48 21 1,184 56 36 23 6.48 22 946 38 26 17 5.66 23 1,004 120 22 21 13.74 24 754 11 27 12 2.94 25 1,802 101 29 25 6.88 26 3,856 52 16 16 1.78 27 747 27 23 19 5.97 28 643 27 19 16 6.50 29 1,130 28 22 17 3.91 Combined 67,804 2,330 1,144 873 4.65 microsatellite-based study, where all these breeds and results revealed that the three different clusters clustered together (Kumar et al., 2006). The high contributed mainly from the Toda, Jaffarabadi, genetic diversity and distinct breed structure imply and Pandharpuri animals, with a very high mem- the possibility of selective breeding in these Indian bership coefficient. The research also stated that buffalo breeds for genetic improvement (Murrah there was an anecdotal evidence to indicate that the and Mehsana). Four breeds (Surti, Pandharpuri, Mehsana breed has been an outcome of gene flow Jaffarabadi and Banni) were distinctly separated from the Murrah males in the recent past. Nili-Ravi while two breeds (Murrah and Mehsana) showed and Murrah have higher average allele frequencies, more admixtures. Admixture was detected in which can be due to biasness to SNP selection from Cluster I of the ancestral clusters, whereas the both Nili-Ravi and Murrah as reference during breeds within remaining clusters were more differ- SNP chip designing. entiated. High admixture was observed between LD decay used to study the linkage of markers Murrah and Mehsana breed, reflecting crossbreed- with increase in intermarker distance and was used ing between these breeds. The probable reason for to decide appropriate intermarker distance for dif- observed admixture in Mehsana could be an out- ferent populations. The magnitude of LD and its come of gene flow from Murrah males in the recent decay with genetic distance determine the resolution past (Kumar et al., 2006) or they might be the same of association mapping and are useful for assessing breed, which was domesticated in different geo- the desired numbers of SNPs on arrays. The results graphical regions. Kumar et al. (2006) evaluated of LD decay illustrate Surti breed showing early the breed admixture using microsatellite markers, decay as compared to other breeds, while Mehsana Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo and Murrah breeds showed late decay together, (mostly of milk protein percentage, milk yield, and which could be assumed as they are under strong milk fat percentage) on chromosome 20 were con- selection pressure. Similar results were obtained by cordant with 13 LD blocks. Mai et al. (2010) rec- Dash et al. (2017) using HD SNP chip for Indian ognized total 98 QTLs for milk production trait, cattle breeds where Sahiwal and Tharparkar breeds which included 30 for milk index, 50 for fat index, showed late decay. These results reflected that the and 18 for protein index. The density of QTLs of Surti breed has smaller population size as it got de- body weight was higher on chromosome 23 along cayed earlier. Other breeds also exhibited LD decay with other productive traits. Mai et al. (2010) also as per their available breedable population. Larger reported a greater number of significant SNP asso- the population size, longer the LD decay. Effective ciations for production (54) than for fertility traits population size of Murrah and Mehsana has dras- (29) with 22 QTL regions associated with fertility tically declined over the last 100 generations. The traits and 14 with production traits. Li et al. (2018) probable cause of drastic decline in Ne for Murrah have used 90K Affymetrix Axiom Buffalo SNP and Mehsana may be attributed to selection efforts Array to identify the SNPs, genomic regions, and done by traditional farmers, as well as the use of genes that were associated with reproductive traits, AI in the native tract of these breeds. It is believed and they have found a total of 40 suggestive loci that Mehsana breed has been developed a couple of (related to 28 genes) that were identified to be asso- decades ago from Murrah and Surti buffalo (might ciated with six reproductive traits (first, second, and have completed less than 100 generations). Hence, third calving age, calving interval, the number of the results should be viewed considering theoret- services per conception, and open days). The con- ical expectations. It gives information regarding cordance study of meat and carcass trait revealed effective population size of ancestors. Shin et al. that the largest QTL of shear force was observed on (2013) estimated the effective population size in chromosome 6 and QTL of tridecylic acid content Korean cattle using HD SNP chip, which revealed located on chromosome 15. Wu et al. (2014) studied rapid increase in effective population size over the carcass trait of Simmental cattle and identified the past 10 generations with the values increasing that the genes in the beef cattle genome signifi- 5-fold (close to 500) by 10 generations. Santana cantly associated with foreshank weight and trigly- et al. (2011) also reported a small effective size of ceride levels. A total of 12 and 7 SNPs in the bovine 40 from several Murrah herds based on phenotypic genome were significantly associated with fore- recordings and average relatedness. An effective shank weight and triglyceride levels, respectively. population size of at least 50 animals is enough In the concordance analysis of exterior traits, to prevent inbreeding depression, the minimum majorly the QTLs were associated with udder traits level recommended by the Food and Agriculture (udder swelling score QTL, udder depth QTL, Organisation of the United Nations. The H level udder attachment QTL, teat length QTL, etc.). is similar in all breeds studied irrespective of their This information of genotypes could be used to population size, but still the present results should associate phenotypes and perform the selection. be interpreted with caution as, for some breeds, less Based on the above results, we can assume that ex- than 50 animals were tested. terior traits are less important for the association The haplotype block structure and its distribu- of QTL with LD block or haplotypes due to the tion in the genome of cattle, especially studies based insufficient size of QTL and low proportion of con- on high-density SNPs, have been rarely reported cordant QTL with LD blocks. van den Berg et al. (Villa-Angulo et al., 2009). However, Bohmanova (2014) studied the concordance for a leg conform- et al. (2010) have performed study on LD for identi- ation trait in dairy cattle and QTL status was used fying the genomic region in American Holstein and in a concordance analysis to reduce the number they have concluded that LD values get inflated by a of candidate mutations. In the concordance study small population and strongly depend on allele fre- of health trait, QTLs associated with somatic cell quency. Thus, the current analysis was performed count were observed almost on every chromosome. to construct the haplotype structure in the buffalo The larger-size QTL of cold tolerance was observed genome and to detect the relevant genes affecting on chromosome 7. More numbers of QTLs asso- quantitative traits. Jiang et al. (2010) identified the ciated with bovine tuberculosis susceptibility were milk trait QTL-specific SNPs in cattle and found found on chromosome 20 and QTLs for clinical that a large proportion of the significant SNPs (61 mastitis found on chromosome 14 as well as on out of 105) were located on BTA14 and also within chromosome 24. Raphaka et al. (2017) identified the reported QTL regions. In our study, 76 QTLs the markers associated with tuberculosis on Bos Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. selection: the comparison of independent samples and taurus autosomes (BTA) 2 and on BTA 23 and con- the identification of regions associated to traits. BMC cluded a major role of BTA 23 for susceptibility to Genomics 10:178. doi:10.1186/1471-2164-10-178. bovine tuberculosis. Barrett, J. C., B. Fry, J. Maller, and M. J. Daly. 2004. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21:263–265. doi:10.1093/bioinformatics/ CONCLUSION bth457. Bohmanova, J., M. Sargolzaei, and F. S. Schenkel. 2010. The study of population structure analysis in Characteristics of linkage disequilibrium in North Indian buffaloes based on SNPs revealed that the American Holsteins. BMC Genomics 11:421. distribution of SNP markers across the buffalo doi:10.1186/1471-2164-11-421. genome of all breeds studied was almost similar. Chang, C. C., C. C. Chow, L. C. Tellier, S. Vattikuti, Genetic parameters, such as H and H , showed S. M. Purcell, and J. J. Lee. 2015. Second-generation E O PLINK: rising to the challenge of larger and richer data- the absence of genetic differentiation while F IS sets. Gigascience 4:7. doi:10.1186/s13742-015-0047-8. was able to differentiate Banni and Surti animals. Corbin, L. J., A. Y. Liu, S. C. Bishop, and J. A. Woolliams. Furthermore, F was able to differentiate the buf- ST 2012. Estimation of historical effective popula- faloes as per their respective geographical loca- tion size using linkage disequilibria with marker tions. The levels of SNPs variation in this study data. J. Anim. Breed. Genet. 129:257–270. doi:10.1111/j.1439-0388.2012.01003.x. could be insufficient to differentiate the other local Dadi, H., J. J. Kim, D. Yoon, and K. S. Kim. 2012. Evaluation breed except Pandharpuri, Surti, and Jaffarabadi of single nucleotide polymorphisms (SNPs) genotyped (morphological distinct breeds), so there is a need by the illumina bovine SNP50K in cattle focusing on to develop SNP chip based on SNP markers iden- Hanwoo breed. Asian-Australas. J. Anim. Sci. 25:28–32. tified by sequence information of local buffalo doi:10.5713/ajas.2011.11232. breeds. LD block QTLs concordance study could Dash, S., A. Singh, A. Bhatia, S. Jayakumar, A. Sharma, S. Singh, I. Ganguly, and S. Dixit. 2017. Evaluation of bo- explore a new window for genomic selection in ani- vine high-density SNP genotyping array in indigenous dairy mals and, therefore, the development of new SNP cattle breeds. Anim. Biotechnol. 29:129–135. doi:10.1080/10 chip based on information of buffalo genome and 495398.2017.1329150. buffalo-specific genetic technologies is warranted. Department of Animal Husbandry, Dairying & Fisheries, Government of India. 2015. Annual report 2015–16. Krishi Bhawan, New Delhi: Department of Animal SUPPLEMENTARY DATA Husbandry, Dairying and Fisheries. Supplementary data are available at Diamond, J. 2002. Evolution, consequences and future of plant and animal domestication. Nature 418:700–707. Translational Animal Science online. doi:10.1038/nature01019. Conflict of interest statement. The authors de- FAO. 2007. Animal genetic resources for food and agriculture. clare that there is no conflict of interest. Rome, Italy: FAO. Fitch, W. M., and E. Margoliash. 1967. Construction of phylogenetic trees. Science 155:279–284. doi:10.1126/ DATA AVAILABILITY science.155.3760.279. The data sets generated and/or analyzed dur- Flori, L., S. Fritz, F. Jaffrézic, M. Boussaha, I. Gut, S. Heath, J. L. Foulley, and M. Gautier. 2009. The genome response ing the current study are not publicly available to artificial selection: a case study in dairy cattle. Plos One [Intellectual Property Rights likely to be generated 4:e6595. doi:10.1371/journal.pone.0006595. out of this and hence the data are not provided Gabriel, S. B., S. F. Schaffner, H. Nguyen, J. M. Moore, J. Roy, publically] but are available from the corresponding B. Blumenstiel, J. Higgins, M. DeFelice, A. Lochner, author on reasonable request. M. Faggart, et al. 2002. The structure of haplotype blocks in the human genome. Science 296:2225–2229. doi:10.1126/science.1069424. LITERATURE CITED Goudet, J. 2005. Hierfstat, a package for R to compute and Anderson, E. C., and J. Novembre. 2003. Finding haplo- test hierarchical F-statistics. Mol. Ecol. Notes. 5:184–186. type block boundaries by using the minimum-descrip- doi:10.1111/j.1471-8286.2004.00828.x. tion-length principle. Am. J. Hum. Genet. 73:336–354. Hu, Z. L., C. A. Park, X. L. Wu, and J. M. Reecy. 2013. Animal doi:10.1086/377106. QTLdb: an improved database tool for livestock animal Barbato, M., P. Orozco-terWengel, M. Tapio, and QTL/association data dissemination in the post-genome M. W. Bruford. 2015. SNeP: a tool to estimate trends era. Nucleic Acids Res. 41(Database issue):D871–D879. in recent effective population size trajectories using doi:10.1093/nar/gks1150. genome-wide SNP data. Front. Genet. 6:109. doi:10.3389/ Iamartino, D., J. L. Williams, T. Sonstegard, J. Reecy, fgene.2015.00109. C. V. Tassell, E. L. Nicolazzi, S. Biffani, F. Biscarini, Barendse, W., B. E. Harrison, R. J. Bunch, M. B. Thomas, and S. Schroeder, and D. A. De Oliveira. 2013. The buf- L. B. Turner. 2009. Genome wide signatures of positive falo genome and the application of genomics in animal Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 SNP-based breed diversity of Indian buffalo management and improvement. Buffalo Bull. 32:151–158. diversity revealed by high-resolution scanning of human doi:10.13140/RG.2.2.30951.85924. chromosome 21. Science. 294:1719–1723. doi:10.1126/ Jiang, L., J. Liu, D. Sun, P. Ma, X. Ding, Y. Yu, and Q. Zhang. science.1065573. 2010. Genome wide association studies for milk produc- Perez-Pardal, L., S. Chen, V. Costa, X. Liu, J. Carvalheira, and tion traits in Chinese Holstein population. PLoS One A. Beja-Pereira. 2018. Genomic differentiation between 5:e13661. doi:10.1371/journal.pone.0013661. swamp and river buffalo using a cattle high-density single Joshi, J., R. K. Salar, P. Banerjee, U. S, M. S. Tantia, and nucleotide polymorphisms panel. Animal 12:464–471. R. K. Vijh. 2013. Genetic variation and phylogenetic re- doi:10.1017/S1751731117001719. lationships of Indian buffaloes of Uttar Pradesh. Asian- Pundir, R., G. Sahana, N. Navani, P. Jain, D. Singh, S. Kumar, Australas. J. Anim. Sci. 26:1229–1236. doi:10.5713/ and A. Dave. 2000. Characterization of Mehsana buffa- ajas.2012.12669. loes in India. J. Anim. Genet. Res. 28:53–62. Joshi, J., R. K. Salar, P. Banerjee, U. Sharma, M. S. Tantia, and Purcell, S., B. Neale, K. Todd-Brown, L. Thomas, M. A. Ferreira, R. K. Vijh. 2015. Assessment of genetic variability and D. Bender, J. Maller, P. Sklar, P. I. de Bakker, M. J. Daly, structuring of riverine buffalo population (Bubalus buba- et al. 2007. PLINK: a tool set for whole-genome associ- lis) of Indo-Gangetic basin. Anim. Biotechnol. 26:148– ation and population-based linkage analyses. Am. J. Hum. 155. doi:10.1080/10495398.2014.955613. Genet. 81:559–575. doi:10.1086/519795 Kataria, R. S., S. Sunder, G. Malik, M. Mukesh, P. Kathiravan, Qin, Z. S., T. Niu, and J. S. Liu. 2002. Partition-ligation– and B. P. Mishra. 2009. Genetic diversity and bottleneck expectation-maximization algorithm for haplotype infer- analysis of Nagpuri buffalo breed of India based on ence with single-nucleotide polymorphisms. Am. J. Hum. microsatellite data. Genetika 45:941–948. doi:10.1134/ Genet. 71:1242–1247 S1022795409070102. Quinlan, A. R., and I. M. Hall. 2010. BEDTools: a flexible suite Kumar, S., J. Gupta, N. Kumar, K. Dikshit, N. Navani, P. Jain, of utilities for comparing genomic features. Bioinformatics and M. Nagarajan. 2006. Genetic variation and relation- 26:841–842. doi:10.1093/bioinformatics/btq033. ships among eight Indian riverine buffalo breeds. Mol. Raj, A., M. Stephens, and J. K. Pritchard. 2014. fastSTRUC- Ecol. 15:593–600. doi:10.1111/j.1365-294X.2006.02837.x. TURE: variational inference of population structure in Kumar, S., M. Nagarajan, J. S. Sandhu, N. Kumar, and large SNP data sets. Genetics 197:573–589. doi:10.1534/ V. Behl. 2007. Phylogeography and domestica- genetics.114.164350. tion of Indian river buffalo. BMC Evol. Biol. 7:186. Raphaka, K., O. Matika, E. Sánchez-Molano, R. Mrode, doi:10.1186/1471-2148-7-186. M. P. Coffey, V. Riggio, E. J. Glass, J. A. Woolliams, Li, J., J. Liu, G. Campanile, G. Plastow, C. Zhang, Z. Wang, S. C. Bishop, and G. Banos. 2017. Genomic regions M. Cassandro, B. Gasparrini, A. Salzano, G. Hua, underlying susceptibility to bovine tuberculosis in et al. 2018. Novel insights into the genetic basis of buf- Holstein-Friesian cattle. BMC Genet. 18:27. doi:10.1186/ falo reproductive performance. BMC Genomics 19:814. s12863-017-0493-7. doi:10.1186/s12864-018-5208-6. Santana, M. L., Jr, R. R. Aspilcueta-Borquis, A. B. Bignardi, MacEachern, S., B. Hayes, J. McEwan, and M. Goddard. 2009. L. G. Albuquerque, and H. Tonhati. 2011. Population An examination of positive selection and changing ef- structure and effects of inbreeding on milk yield and fective population size in Angus and Holstein cattle popu- quality of Murrah buffaloes. J. Dairy Sci. 94:5204–5211. lations (Bos taurus) using a high density SNP genotyping doi:10.3168/jds.2011-4377. platform and the contribution of ancient polymorphism Shah, T. M., J. S. Patel, C. D. Bhong, A. Doiphode, to genomic diversity in Domestic cattle. BMC Genomics U. D. Umrikar, S. S. Parmar, D. N. Rank, J. V. Solanki, 10:181. doi:10.1186/1471-2164-10-181. and C. G. Joshi. 2013. Evaluation of genetic diversity and Machado, M. A., I. Schuster, M. L. Martinez, and population structure of west-central Indian cattle breeds. A. L. Campos. 2003. Genetic diversity of four cattle Anim. Genet. 44:442–445. breeds using microsatellite markers. Rev. Bras. Zootecn. Shin, D. H., K. H. Cho, K. D. Park, H. J. Lee, and H. Kim. 32:93–98. doi:10.1590/S1516-35982003000100012. 2013. Accurate estimation of effective population size Mai, M. D., G. Sahana, F. B. Christiansen, and in the Korean dairy cattle based on linkage disequilib- B. Guldbrandtsen. 2010. A genome-wide association study rium corrected by genomic relationship matrix. Asian- for milk production traits in Danish Jersey cattle using a Australas. J. Anim. Sci. 26:1672–1679. doi:10.5713/ 50K single nucleotide polymorphism chip. J. Anim. Sci. ajas.2013.13320. 88:3522–3528. doi:10.2527/jas.2009-2713. Sraphet, S., B. Moolmuang, A. Na-Chiangmai, S. Panyim, Nagarajan, M., K. Nimisha, and S. Kumar. 2015. D. R. Smith, and K. Triwitayakorn. 2008. Use of cattle Mitochondrial DNA variability of domestic river buffalo microsatellite markers to assess genetic diversity of (Bubalus bubalis) populations: genetic evidence for domes- Thai Swamp buffalo (Bubalus bubalis). Asian-Austral. tication of river buffalo in Indian Subcontinent. Genome J. Anim. Sci. 21:177–180. doi:10.5713/ajas.2008.70327. Biol. Evol. 7:1252–1259. doi:10.1093/gbe/evv067. Suh, S., Y. S. Kim, C. Y. Cho, M. J. Byun, S. B. Choi, Y. G. Ko, Nayee, N. G., R. M. Ambaliya, S. Saha, K. R. Trivedi, and C. W. Lee, K. S. Jung, K. H. Bae, and J. H. Kim. 2014. M. N. Namjoshi. 2016. National database for enhancing Assessment of genetic diversity, relationships and struc- the productivity of cattle and buffaloes—suitability of ture among Korean native cattle breeds using microsatel- INAPH to create a national database. Indian Dairyman lite markers. Asian-Australas. J. Anim. Sci. 27:1548–1553. 68(9):84–88. doi:10.5713/ajas.2014.14435 Patil, N., A. J. Berno, D. A. Hinds, W. A. Barrett, J. M. Doshi, Sved, J. A. 1971. Linkage disequilibrium and homozygosity of C. R. Hacker, C. R. Kautzer, D. H. Lee, C. Marjoribanks, chromosome segments in finite populations. Theor. Popul. and D. P. Mcdonough. 2001. Blocks of limited haplotype Biol. 2:125–141. doi:10.1016/0040-5809(71)90011-6. Translate basic science to industry innovation Downloaded from https://academic.oup.com/tas/article/5/2/txab033/6271363 by guest on 11 May 2021 Thakor et al. Tantia, M., R. Vijh, B. Mishra, S. Kumar, and R. Arora. Simmental cattle. PLoS One 9:e109330. doi:10.1371/ 2006. Multilocus genotyping to study population journal.pone.0109330. structure in three buffalo populations of India. Asian- Zhang, K., M. Deng, T. Chen, M. S. Waterman, and F. Sun. Austral. J. Anim. Sci. 19:1071–1078. doi:10.5713/ 2002. A dynamic programming algorithm for haplotype ajas.2006.1071. block partitioning. Proc. Natl. Acad. Sci. USA 99:7335– Toro, M., and A. Maki-Tanila. 2007. Genomics reveals do- 7339. doi:10.1073/pnas.102186799. mestication history and facilitates breed development. In: Zhang, K., F. Sun, M. S. Waterman, and T. Chen. 2003. K. Oldenbroek, editor, Utilization and conservation of Dynamic programming algorithms for haplotype block farm animal genetic resources. Wageningen, Netherlands: partitioning: applications to human chromosome 21 Wageningen Academic Publishers. p. 75–102. haplotype data. In: Vingron, M., S. Istrail, P. A. Pevzner, van den Berg, I., S. Fritz, S. Rodriguez, D. Rocha, and M. S. Waterman, editors. Proceedings of the Seventh M. Boussaha, M. S. Lund, and D. Boichard. 2014. Annual International Conference on Research in Concordance analysis for QTL detection in dairy cattle: Computational Molecular Biology. New York, NY, USA: a case study of leg morphology. Genet. Sel. Evol. 46:31. Association for Computing Machinery; p. 332–340. doi:10.1186/1297-9686-46-31. Zhao, H., D. Nettleton, and J. C. Dekkers. 2007. Evaluation Villa-Angulo, R., L. K. Matukumalli, C. A. Gill, J. Choi, of linkage disequilibrium measures between multi-allelic C. P. Van Tassell, and J. J. Grefenstette. 2009. High- markers as predictors of linkage disequilibrium between resolution haplotype block structure in the cattle genome. single nucleotide polymorphisms. Genet. Res. 89:1–6. BMC Genet. 10:19. doi:10.1186/1471-2156-10-19. doi:10.1017/S0016672307008634. Wu, Y., H. Fan, Y. Wang, L. Zhang, X. Gao, Y. Chen, Zhivotovsky, L. A. 2015. Relationships between Wright’s F ST J. Li, H. Ren, and H. Gao. 2014. Genome-wide associ- and F statistics in a context of Wahlund effect. J. Hered. IS ation studies using haplotypes and individual SNPs in 106:306–309. doi:10.1093/jhered/esv019. Translate basic science to industry innovation
Journal
Translational Animal Science – Oxford University Press
Published: Apr 1, 2021