CandiHap: a toolkit for haplotype analysis for sequence of samples and fast identification of candidate causal gene(s) in genome-wide association study

Abstract

Genome-wide association study (GWAS) is widely used to identify genes involved in plants, animals and human complex traits. Generally, the identified SNP is not necessarily the causal variant, but it is rather in linkage disequilibrium (LD). One key challenge for GWAS results interpretation is to rapidly identify causal genes and provide profound evidence on how they affect the trait. Researches want to identify candidate causal variants from the most significant SNPs of GWAS in any species and on their local computer, while to complete these tasks are to be time-consuming, laborious and prone to errors and omission. To our knowledge, so far there is no tool available to solve the challenge for GWAS data very quickly. Based on the standard VCF (variant call format) format, CandiHap is developed to fast preselection candidate causal SNPs and gene(s) from GWAS by integrating LD result, SNP annotation, haplotype analysis and traits statistics of haplotypes.

Multi-omics analyses of 398 foxtail millet accessions reveal genomic regions associated with domestication, metabolite traits, and anti-inflammatory effects

“…Given the false-positive and false-negative issues associated with GWAS (Atwell et al, 2010;Chan et al, 2010aChan et al, , 2010b, we combined biological and bioinformatic approaches to mine candidate genes. The filtering procedure was set as follows: (1) haplotype analysis for all genes in the LD regions of a significant SNP using CandiHap (Li et al, 2020) (Supplemental Figure 4); (2) search for proteins that are biochemically related to the associated metabolic traits encoded at the candidate loci; (3) cluster analysis of candidate genes and homologous genes with known function; (4) cross-reference with statistical significance results of causal genes from gene-based GWAS using CandiHap; and (5) prediction of functional/molecular phenotypes from GWAS results by performing transcriptome-wide association studies (TWASs) (Supplemental Table 12). After the filtering steps above, we identified 511 candidate genes associated with 690 lead SNPs that might be responsible for the variation in metabolic traits (Table 1 and Supplemental Table 13).…”

“…To the best of our knowledge, haplotype analysis has usually been conducted manually, which is laborious, time-consuming, and prone to errors and omissions. To solve this problem, we developed a user-friendly software, CandiHap (https://github.com/xukaili/CandiHap) (Li et al, 2020). With CandiHap, users can complete the following analysis within a minute, which usually costs hours or even days manually: (1) local GWAS for a gene; (2) haplotype analysis for a gene (CandiHap); (3) haplotype analysis for all genes in the LD regions of a significant SNP position one by one (GWAS_LD2haplotypes); (4) haplotype analysis of Sanger sequencing population variation data run on Unix-like systems (sanger_CandiHap.sh).…”

Population Genomics and Haplotype Analysis in Bread Wheat Identify a Gene Regulating Glume Pubescence

“…A total of 52 SNPs on the gene sequence of TaELD1-1A among the 352 wheat accessions were retrieved from the 1000 wheat exomes project of He et al (2019) ( Supplementary Table 5 ), among which 14 SNPs were filtered with heterozygosity <0.03 and used for haplotype analysis by the “CandiHap” package ( Li et al, 2020 ) of R 4.0.1 ( R Core Team, 2013 ) (see text footnote 4) ( Supplementary Table 6 ). Haplotype analysis showed that four main haplotypes (Hap1–4, containing accessions >10) of TaELD1-1A were detected among 352 wheat accessions ( Figure 4 and Supplementary Table 7 ).…”

“…To assess the allelic variation of the TaELD1-1A gene across various wheat cultivars, the haplotype analysis of TaELD1-1A was performed using the SNP data (heterozygosity <0.03) on TaELD1-1A gene sequences among the 352 wheat accessions retrieved from the 1000 wheat exomes project of He et al (2019) (see text footnote 3) using the “CandiHap” package ( Li et al, 2020 ) of R 4.0.1 (R Core Team, 2013) (see text footnote 4), and the differences of the phenotypes for Gp corresponding to different haplotypes were tested. Moreover, the homologous gene sequences of TaELD1-1A in pan-genomes including 10+ hexaploid wheat ( Walkowiak et al, 2020 ), emmer wheat (Zavitan) ( Avni et al, 2017 ), and durum wheat (Svevo) ( Maccaferri et al, 2019 ) genomes were downloaded from Ensembl Plants 8 according to the best-match gene IDs to TraesCSU02G143200 through BLAST.…”

Genetic basis of the early heading of high-latitude weedy rice

“…The genes located within the high LD blocks were selected for checking the gene annotation (https:// rapdb.dna.affrc.go.jp/). Then, a haplotype-based association analysis for the genes within the LD block was conducted by candihap software (Li et al, 2020).…”