We are working on a project where we need to perform genome wide association studies for multi-parental population. We have the SNP data from whole genome re-sequencing protocol. As I am quite new to this field, I would like what are the filtering criteria one need to employ before moving towards finding association study.
Thanks in advance.
The most important aspect of SNP calling from genomic data is depth. Make sure the SNPs are called at higher depth at least more than 30X.
Hi Sourav,
Let me know if you have reference lines and in which crop you are working to simplify your problem?
Since you are going for GWAS with multi parents....
1. Look for the SNP's which are polymorphic between the parents.
2. Avoid the the SNP's within 50 bps range.
3. Eliminate the complementary SNP's (A-T,G-C) vice versa.
4. work for the functional SNP's checking in the public data base which will be more use full.
5. Finally try to cover the whole genome with equal no of SNP's distribution thought the genome.
Good luck.
Dear all,
Thank you for your valuable suggestions. These actually helped me a lot.
@ Anand: I am working on Chickpea and we have also sequenced the reference lines.
We will be performing GWAS only on RIL population which is derived from multi-parents. Thus is it necessary to perform filtering with respect to parents?
Thanks again.
Good to know that you are working on Chickpea...
Yes, It is necessary to look polymorphic SNP's between the parents from which the RILs were developed.
Applying this filter the SNP's which you design all of them will be useful for the experiment.
Good luck
Hi Marco;
Here are some quality checking in GWAS;
- In general, samples reflecting duplicates, low call rates, gender mismatch, or population outliers are excluded.
- Variants are excluded by the following criteria
- call rate < 0.95,
- minor allele count (MAC) 6,
- Hardy-Weinberg Equilibrium (HWE) P < 1×10−4 ,
- If you have imputing results then imputation quality score < 0.3 for minimac or < 0.4 for IMPUTE, or absolute allele frequency difference > 0.3 compared with expected allele frequency
Hi Sourav,
Great question. Are you talking about DNA microarray data, targeted array (e.g. whole exome) or whole genome sequencing data?
The answer to your question varies a bit depending on the technological platform used.
Dear all,
Sorry for writing back back late.
@ Vincent: We have performed low coverage (4-5X) whole genome re-sequencing for a RIL population consisting more than 1000 genotypes.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics