II performed SNP analysis on a streptomyces species strain. The different strains were selected over several generations. However, only 20-25% of reads show an alternate allele. During SNP analysis of cancer cells, that would be normal because of the tumor heterogeneity. However, I did not expect that in streptomyces. Streptomyces do show mycelial growth, one possibility is that the cells never became single cells during selection leading to the heterogeneity. Is there any other explanation?
This could have several problems:
1. Which mapper are you using?
2. Which SNP caller + parameter.
3. Is there a high qual ref. sequence
5. How deep is the coverage?
Hi,
Yes, a single cell dilution method would be good to get a homogenous population, Try to get a single cell of the aerial mycelium if possible.
This article might be of some help.....
https://www.jic.ac.uk/staff/mark-buttner/pdfs/Flardh%20and%20Buttner%202009.pdf
Best Wishes
Dilantha
Thanks for all the responses...
Hello Fritz, I am using the Basic Variant detection program of CLC Genomics Workbench.
The parameters I set include a minimum of 20 reads in the region with minimum 20% variant population and the reference sequence is of pretty high quality.
8.4 million reads of 250bp length were used for the mapping to a reference of 9.5 mega bases, so the coverage is pretty good.
Hello Rohan and Dilantha,
These strains were generated long back and used for years before being sequenced. All this and the sequencing was done before I arrived at the scene. The people who worked with the strains at that time tried to maintain homogeneity as much as they could but it is possible it was not completely homogeneous. There is nothing that can be done about the strain or sequencing now, so is there anything other than heterogeneity that can explain my observation?
Hello Jun Cui,
It is a diploid organism, that's why I am surprised at 20-25% of reads showing alternate allele and other than heterogeneity, I can't think of any other explanation.
Hi,
so how many SNPs over the whole genome do you find? This is the number that is more intuitive then the 20% of reads... Because it could be that all 20% lie in one region :).
You could also try to manual assess some of those snps by using eg. IGV.
I am still not sure what you actually expected. If there are "too view SNPs" then you could try a different mapper/SNP caller or simply parameters.
If you expected no SNPs, I can tell you that this never happens. The reference sequences are always different from the organism that is sequenced. There are mutations happening in every cell (e.g. somatic mutations) . As you said, those strains are living for ages in your lab so there will be differences.
Hope that helps
Hello Fritz,
I got a total of 971 alterations (mostly SNPs but a few MNVs and Indels). Of them 382 have an average PHRED score of 20 or higher. The total number of SNPs is on expected lines as these were subjected to chemical mutagenesis. However, in some of these SNPs, only about 20-30% of the reads show the altered allele. Given that the organism is diploid, the only way I can explain only 20-30% of the reads showing mutation is heterogeneity. Is there anything else that can explain it?
Hello Arkadeep,
It seems that you thought you would have a higher percentage of alternative alleles during your analysis, but what was your expected minor allele frequency (MAF)? I know in humans MAFs can range from 1% (or lower) to 50%; a MAF lower than 50% can be stable in a population if the alternative allele does not result in a biologically significant impact on the organism. If the 20% to 25% your reading is significantly lower than the expected MAF then it means there is likely selection pressure (or as you already pointed out you're analysing heterogenous populations).
Hope this helps,
Lance
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics