Skip to content

ttchen0714/BRCAness-signature

Repository files navigation

BRCAness signature Manual

1. Within-sample relative expression orderings (REO)

Within-sample REO was defined as a binary relation between gene i and gene j, where Gi > Gj or Gi < Gj. Gi and Gj denote the expression values of gene i and gene j, respectively. A set of REOs is highly stable in OvCa with BRCA1/2 mutation and reversed in patients with wild type BRCA1/2, which reflects the feature of BRCAness.



2. Training the BRCAness signature (2-GPS)

The workflow of developing the REO-based BRCAness signature:
(1) we constructed a discrete profile based on the REOs of gene pairs;
(2) we defined the reversed gene pair with a stable REO pattern (Gi > Gj) in more than 99% of mutation samples, but with a significantly reversed REO pattern (Gi < Gj) in wild-type samples tested by Fisher’s test with p < 0.05;
(3) a greedy algorithm was used to generate a set of gene pairs with high coverage in wild-type samples with iteration conditions. First, the gene pair with the highest coverage was chosen as the seed. Second, a candidate gene pair was taken into account that increases the most coverage and more than 1%.



Finally, using the rank-based algorithm, we discovered a BRCAness signature consisting of 2 gene pairs, including CXCL1 > SV2A and LY9 > CHRNB3, using gene expression profiles of ovarian cancer from the Cancer Genome Atlas.


3. Application

RDtij=Rti-Rtj;

Where RDtij reprents the rank difference between gene i and gene j in sample t, Rti reprents the rank of gene i in sample t and Rtj reprents the rank of gene j in sample t.

BRCAness score(t)=count(RDtij>0)/N.  

For the OvCa sample t, the BRCAness score was calculated as the proportion of gene pairs (Gi > Gj) with values of rank differences more than 0 among the total gene pairs of the BRCAness signature (N=2). According to a determination rule, samples were divided into "BRCAness group" (BRCAness score = 1) or "non-BRCAness group" (BRCAness score < 1).

About

BRCAness-signature

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages