search

mskcc / facets

Algorithm to implement Fraction and Copy number Estimate from Tumor/normal Sequencing.
144 stars 67 forks source link

through emcncf(oo)$cncf to annotate concrete genes #120

Closed kobejamescurry closed 5 years ago

kobejamescurry commented 5 years ago

Sorry, I'm bombarding you with questions... I'd like to make sure my understanding is correct. dear professor, thanks for your so powerful tool. I have a annotation gene question with the result. image

and I download a annotate file from ucsc and extracted it as a bed.format, and I extract chr start end from the emcncf(oo)$cncf, just like following image

and a annotate bed like following image then use bedtools intersect -a f4.bed -b all.bed -wb > f5.bed to get the result.

image So all the gene annotated , combining the tcn and lcn result to define LOH

For example , all gene in chr 2116996 16775632 can be defined as LOH, is that right? thanks a lot

kobejamescurry commented 5 years ago

The raw total copy number for a segment is 2*2^(cnlr.median - dipLogR). This is attenuated by the purity. For example if the copy number is 3 and purity is 70% you should get 2.7. However, since it's an estimate it won't be exactly 2.7. The extent of error depends on the number of markers in the segment among other factors (such as dipLogR and purity estimates which best fit the whole genome and not just the segment).

and I think tcn and lcn is all integer, it maybe not normal?