search

statgen / demuxlet

Genetic multiplexing of barcoded single cell RNA-seq
Apache License 2.0
121 stars 26 forks source link

demuxlet when genotype data is not available for one of the samples in the pool #71

Closed ashmilrp closed 4 years ago

ashmilrp commented 4 years ago

Hello,

I am using demuxlet to deconvolute sample identity and have a few questions about the tool:

  1. Is it recommended to run demuxlet when 5 samples are pooled by barcoded single cell sequencing and genotype data of 1 of the 5 samples is not available. If so how should the parameters be set (for example, should the alpha be set to 0.2?) ?

  2. I noticed demuxlet is also available in popscle: https://github.com/statgen/popscle. Is there any major change applied in version2 that will affect the results?

  3. I am trying to understand the difference between DBL-293T_RTG-jurkat-0.500 and DBL-jurkat-293T_RTG-0.500. Is this related to the sample order in the genotype VCF? When I ran demuxlet on the example dataset with default parameters I see the best assignment for 6 barcodes as DBL-jurkat-293T_RTG-0.500 while all others are assigned as singlets. But when the order of samples are switched in the jurkat_293t_exons_only.vcf.gz there are 5 barcodes has best assignment to DBL-293T_RTG-jurkat-0.500 and 1 has DBL-jurkat-293T_RTG-0.500.

It would be very helpful for us to better infer the demuxlet results if we have answers to the above questions. Please let me know if you need more information from my side.

Thanks in advance for your time and effort!

Best, Ashmitaa