I'm developing an R-based workflow using crlmm, which uses a normalization method different from what Illumina's proprietary BeadStudio software performs. I'm not sure exactly how it differs, and is it better or worse for my purposes? Specifically, I am starting with paired Grn/Red IDAT files for tumor-normal samples per patient. After probe-level copy-number estimation using crlmm, I'll do segmentation of the probes (using CBS), and identify somatic CNVs (subtract normal CNAs from the tumor). Those somatic CNVs will be fed into GISTIC to find recurrently/focally altered genes.
Any insight into my proposed workflow would be massively helpful. I'm sure that there might be a better solution. I would like to avoid using any proprietary software since this needs to be automated from start to end.
Cheers!
~Cyriac
0 answers
No answers yet.
Log in to answer this question.
-
PureCN issue: having problems running a small example
written by bboy •
-
Using GRIN on WES data and general driver analysis preprocessing
written by andrei.rajkovic •
-
Tutorial: Tutorial: Analyze exome Copy number variation (CNV) in single patient or in population.
written by Chirag Nepal
-
Tutorial: Working with MAF files (Mutation Annotation Format) from the TCGA (The Cancer Genome Atla…
written by Cyriac Kandoth
Don't have experience with those packages, but given that you have matched tumor/normal pairs, the normalization matters a bit less - some of that noise will cancel out when you calculate the difference between the samples.