Hi, we have two Illumina-sequenced plant samples where one is a mutant of the other. Do you know a straightforward method to compare these in order to spot small (SNPs, InDels) or maybe longer mutations (LTR insertion, don't think CNV) or do you suggest align both to a reference, call mutations and then compare?
Tnx
Hello, I think it also depends a bit on the hardware you have on your disposal. For genome comparison I can recommend LAST (http://last.cbrc.jp/) with which I achieved good results. Should also be fairly fast. But wouldn't a simple global alignment between both genomes do the trick? I guess for other Variant Calling tools you need more of a population to perform the analysis.
Dear Marco,
You may consider artMAP tool for your purposes (https://www.biorxiv.org/content/early/2018/09/12/414433.full.pdf+html). The final analysis of the tool will only give snp having g to a and c to t conversion as EMS-induced those. However, the pipeline will produce VCF files from both samples with SNPs, and you can use bedtool subtract function or any tool you like to get mutant-specific SNPs easily. Also, the tool is ready-to-use (Linux/Mac/Win) since it is provided through a Docker container, see here: https://github.com/RihaLab/artMAP/blob/master/artMAP.md.
Otherwise, you can have a look at the galaxy project (https://usegalaxy.org/), where you can set up your own pipeline of analysis or load a pre-defined ones.
I hope this can help you,
Best regards,
Andrea
You can actually do both ways. When NGS first became popular, the only ways to really identify mutations was using a "subtraction" method, where you call mutations in each sample separately and then subtract one from the other to find the differences. This isn't very intuitive though so now there are many more algorithms that will compare the samples directly against each other in a pairwise manner to give you the mutations that are specific to one.
VarScan2, SomaticSniper and MuTect are the ones I have experience in that do this direct comparison. I prefer VarScan2 because it is easy to use and you can do CNV analysis with it.
Article VarScan 2: Somatic mutation and copy number alteration disco...
You align each sample into the reference genome. Generate Variant file each sample. bcftools isec option will help you to split common variant between the sample(wild type and its mutant) and Real mutant variant.
Another way, First you align your wild-type sample into the reference genome and generate the vcf file. Create new consensus from with the help of VCF file. Then align the mutant sample into a new consensus/reference sequence.
SNPtrack may help you.
http://genetics.bwh.harvard.edu/snptrack/
You will try Sophie Sneddon mentioned tools also.
- Badges
- Science method