search

broadinstitute / ichorCNA

Estimating tumor fraction in cell-free DNA from ultra-low-pass whole genome sequencing.
GNU General Public License v3.0
160 stars 87 forks source link

Number of samples for custom PON #81

Open shahmj opened 3 years ago

shahmj commented 3 years ago

Hi,

How many samples do you typically recommend one should use to create a custom PON? I'm trying to create one for GRCh38 with mean cvg around 30x/sample.

Thanks, Minita

aoumess commented 2 years ago

Hi ! I am having the same question (also trying to create a PoN for hg38 for higher resolution than the 500Kb / 1Mb provided with the software). I currently tested building a PoN from 15 randomly chosen WES profiles from the 1000 genomes project, downsampled to 1X using samtools. While I'm sure mimicking a shWGS by artificially downsampling a bulk WGS is wrong in many ways, that's the only public data I've found so far... Results are poor, not that different than processing my samples without a reference at all, unfortunately... I did not find any information on how the provided PoNs were constituted. Could we have the reference for the source of samples, so that we can reproduce them and locally generate their equivalent at higher resolutions, please ? Thanks !