How to check if a sample has been mixed with another sample based on NGS sequencing data

Question

[katjanjarosz]

Hi. I suspect that the sample has been contaminated with another sample. Do you know how to check coverage between one sequenced data with another and see if one is in the other? Thanks

[WouterDeCoster]

how to check coverage between one sequenced data with another and see if one is in the other

Why the coverage? Unless you are expecting contamination from another species, the only way is to look at called variants.

[GenoMax]

Are the samples from the same genome or different? If they are from the same genome it would be impossible to check contamination/coverage unless you had UMI (unique molecular indexes) on your reads. If samples are from different genomes then how different are they (e.g. yeast/human)?

[katjanjarosz]

I have sequencing data from multiple strains of Saccharomyces cerevisiae. We think that one strain contaminated another (both strains were sequenced).

[WouterDeCoster]

Then you need to see if there are variants from your suspected contaminant showing up with low allele frequencies in the sample which you suspect is contaminated.

[Asaf]

I would take the experiment and PCA plot it, if this one sample looks weird I'll remove it.

Answer 1

[tshtatland]

Use variant allele fractions to detect contamination. See, for example:
Same-Species Contamination Detection with Variant Calling Information from Next Generation Sequencing. Tao Jiang, Martin Buchkovich, Alison Motsinger-Reif, bioRxiv 531558; doi: https://doi.org/10.1101/531558
Posted January 26, 2019.
https://www.biorxiv.org/content/10.1101/531558v1