Bulk Segregant Analysis Q&A

General Questions

In BSA (bulked segregant analysis), individuals in a segregating population are grouped and mixed according to the phenotype of the target trait, and the individuals or strains in the population are divided into two groups according to the relative differences of the target trait, and then the DNA of the individuals or strains in the two groups are mixed separately to form a relative DNA mixing pool. The parental and offspring pools are sequenced, SNPs are detected, and the index values of the offspring pools are calculated based on the loci with pure differences between the parents, and the loci with larger differences are selected.

It is recommended to select parents with single trait differences and few heterozygous loci as much as possible, and it is also recommended that the differences between parents should not be too large, because if the parental differences are too large, it will lead to heavy false positives and easy to locate the false positive region; if you select two varieties that are directly related, it is easy to locate the candidate region because the parental differences are too small; for the parents of the target trait, there are various ways to obtain the parents. For the parents of the target trait, there are various methods to obtain the candidate genes, such as EMS mutagenesis, natural individuals, UV mutagenesis, all of which can be used to locate the candidate genes by BSA.

(1) If the trait under study is an EMS-induced mutation trait, only two subgenerational pools (wild type + mutant) can be tested, or one parent (wild type) and one subgenerational pool (mutant) can be tested;
(2) If the trait under study is a quantitative trait, it is recommended to measure two parents and two subgenerational pools. The effect of the number of samples on the analysis results was evaluated by the actual online project, in which the best results were obtained with four samples (i.e., two parents + two subgenerational pools), followed by one parent and two subgenerational pools, and if only two subgenerational pools were measured, the number of false positives would increase and many SNPs would be obtained.

(1) Validation of SNPs. (A) Convert the candidate SNPs into CAPS or dCAPS markers, i.e., perform restriction endonuclease recognition site analysis on the candidate SNPs, screen out the SNPs that cause changes in the enzymatic recognition sites, amplify the fragments where these SNPs are located using the corresponding primers, and then perform enzymatic digestion and electrophoresis on the amplified products to convert the SNPs into CAPS markers; perform polymorphism analysis on the CAPS markers to verify the availability of the markers; (B) The candidate SNPs were amplified by PCR, and then the amplification products were verified by Sanger sequencing;
(2) Validation of candidate gene expression across phenotypes using RT-PCR
(3) Transcriptome-based differential expression analysis to see if there are significant differences in gene expression
(4) RNAi analysis: using RNAi technology to specifically knock out or turn off the expression of specific genes.

There are many factors that lead to unsatisfactory localization, mainly the following:
(1) large differences in genetic background between parents, there are many other differences in addition to the target trait, which creates a lot of interference in the analysis and makes it difficult to localize;
(2) complex trait statistics, the target trait may be composed of several simple traits, which can be split and re-localized, and also the quantitative traits themselves are difficult to locate and have some uncontrollability;
(3) the sequencing data are contaminated, and the comparison results can be checked by extracting part of the sequencing data to do blast comparison in the nr library;
(4) the analysis method is not applicable, we can use ED method and SNP-index method for localization respectively and compare the localization results.

The ideal state is to extract DNA from each daughter sample individually, and then mix them equally and evenly into one DNA according to the concentration of DNA, so that the amount of DNA information of each daughter is equal. However, due to the current low transaction price of sequencing projects, the general company practice is to let customers take equal amounts of tissue mix from each sample and then extract as a mixed pool DNA sample.
Mixing and extracting equal tissue samples is not as effective as mixing after single sample extraction, but its impact is perfectly acceptable. Although mixing first and then extracting will result in unequal amounts of DNA from each sample, if the sampling is ideal enough, the genotypes of these samples at the target loci should be consistent, and the main impact is the composition of the background noise, which does not affect our ability to find the target loci later in the data analysis.