Question: how to decide whether a mutation is driver for a specific cell line
gravatar for wenyu
8 months ago by
wenyu0 wrote:

Hi, I am working on the CCLE cell line mutation data. The mutation data is a summarized list, with each line representing a specific mutation in specific cell as well as other annotations including:

Hugo_Symbol Entrez_Gene_Id NCBI_Build Chromosome Start_position End_position Strand Variant_Classification Variant_Type Reference_Allele Tumor_Seq_Allele1 dbSNP_RS dbSNP_Val_Status Genome_Change Annotation_Transcript Tumor_Sample_Barcode cDNA_Change Codon_Change Protein_Change isDeleterious isTCGAhotspot TCGAhsCnt isCOSMIChotspot COSMIChsCnt ExAC_AF WES_AC WGS_AC SangerWES_AC SangerRecalibWES_AC RNAseq_AC HC_AC RD_AC

now I want to give a binary value or a probability score to denote whether a mutation is driver in this specific cell. I have checked many post and papers, but their method are all based on original sequencing data rather that such a mutation list. Any idea about how to do this? Many thanks in advance!

mutation gene • 364 views
ADD COMMENTlink modified 8 months ago by poisonAlien2.5k • written 8 months ago by wenyu0
gravatar for poisonAlien
8 months ago by
poisonAlien2.5k wrote:


Whether a gene is a driver or not is usually done on large cohorts based on how frequently the gene is mutated. There are algorithms to identify such genes. How many cell lines do you have in your file ?

In case if you're only interested in a particular mutation,

  • you will have to look at literature for its functional studies or
  • look at other factors such as whether its a deleterious mutation, SIFT and PolyPhen scores associated with them, how frequently its mutated in cancers (TCGA cohorts), etc.
  • You could also search for various databases such as CIViC to see if its clinically associated.
ADD COMMENTlink written 8 months ago by poisonAlien2.5k

Hi, currently I am interested in 42 cell lines with each cell line harbouring several hundred or over one thousand mutations, all of them located in coding region and germline mutations were filtered. I understand that most germline mutations are not driver mutations, but all those mutations in my list I am wondering how to decide if they are driver or not.

I am new to this area so please point out if I am wrong. My understand is that a gene harboring driver mutations is a driver gene, but a driver gene can also harbor passenger mutations. Also a mutation being "driver" in one cell doesn't mean it is also driver in other cell. When you say "how frequently the gene is mutated", do you mean "how many cancer patients carrying the mutation"? So technically if a mutation is carried only by all the cancer patients but not by the individuals from healthy cohort, then we can call this mutation a driver?

ADD REPLYlink written 8 months ago by wenyu0
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 2093 users visited in the last hour