Mutation in cancers

Cancer mutations

• Cancer genome instability -> many genome alterations
  • Drivers
    • cell growth advantage
  • Passengers
    • occur by chance
• Recurrent mutations
  • Source
    • “sequencing or mutation caller error” e.g. INDELS(false positives)
    • “high mutation rate” e.g. low expression levels
    • “growth advantage” -> driver

Goals

We’d like to learn

• Distribution
  
• relationship with
  • expression
    
  • cancer type
    
  • etc
• effects

Distribution

Q. Are recurrent mutations enriched in cancer genes? what about POGs and TSGs
A. Yes they are.

Distribution of different gene types

	1<#<5	>5	percent_gt5	enrich
NON_CG	261137	2999	0.0113540	1.000000
POG	5817	205	0.0340418	2.998225
TSG	6891	180	0.0254561	2.242037

*P_value for enrichment are <2.2e-16 by fisher’s exact test

relationship with expression

• Anyone has concept of low expression can be driver?
  No preference found for Cancer genes

Distribution of different gene types in 5 tiles of expression

	1	2	3	4	5	Pvalue
CG	51	51	36	57	34	0.0596961
NON_CG	3423	3399	3428	3408	3439	0.9899763
POG	41	58	62	50	52	0.2947908
POGTSG	10	14	8	7	12	0.5224011
TSG	46	49	37	49	34	0.3303183

Relationship with expression (Frameshift ins/del in_frame ins/del)

• Are recurrent variants more like be high or lower expression for tumor suppressors or oncogenes?
  
  

  Figure.1 number of recurrent mutations in gene as a function of expression level

cont. (Missense, nonsense and silent)

*every point is a gene

Mapping mutations on proteins

• Cbioportal has this feature
  • But they excluded “silent mutations”, unable to customize
• So I did this…
  
• Mapping different mutations onto TP53

Problem

• pfam downloads may fail if too many downloads attempted;