search

broadinstitute / ichorCNA

Estimating tumor fraction in cell-free DNA from ultra-low-pass whole genome sequencing.
GNU General Public License v3.0
164 stars 88 forks source link

ichorCNA copy number ratio in the cna.seg file differs from the genome wide plot #20

Open manasa-surakala opened 6 years ago

manasa-surakala commented 6 years ago

Hi, When there is no normal wig supplied, the minimum log2 ratio is going upto -6 in cna.seg file but that value is not showing up in the plots. Is there any limitation in the plots shown (-2 to 2) ? Also, when I run including a normal wig, then the tumor fractions change a lot. Eg. for one sample it changed from 0.3 to 0. Please comment on this. Thanks, Manasa

gavinha commented 6 years ago

Hi,

Are you using ichorCNA for whole genome or whole exome sequencing data? If it's exome, then you will definitely need a matched normal. This can have a major effect on your results. If it's genome, then you a normal is not required. However, if you do use a matched normal, you should expect cleaner signals and profiles. Is this the case? Do you have examples you are willing to share? You can also send to me privately at gavinha@broadinstitute.org.

You can change the plotting y-axis limits by using --plotYLim "c(-6,6)" when running runIchorCNA.R. If you are using the snakemake pipeline then modify this line https://github.com/broadinstitute/ichorCNA/blob/05519bf02d4ae6c949b89551152cb107f86fdcb2/scripts/snakemake/config/config.yaml#L42