https://github.com/databio/fused
FUSion Error prone repair Detection
https://github.com/databio/fused
Last synced: 2 months ago
JSON representation
FUSion Error prone repair Detection
- Host: GitHub
- URL: https://github.com/databio/fused
- Owner: databio
- License: bsd-2-clause
- Created: 2022-01-11T15:14:56.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2022-05-05T15:40:06.000Z (about 4 years ago)
- Last Synced: 2025-09-11T10:16:14.386Z (10 months ago)
- Language: R
- Size: 113 KB
- Stars: 2
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
**FUS**ion **E**rror-prone repair **D**etection
## Description
FUSED is a collection of functions to determine the presence of hallmarks of common DNA repair mechanisms including NHEJ, MMEJ, SSA, BIR, and MMBIR.
### NHEJ
For non-homologous end-joining, FUSED looks for homology between 1-4nt (nt==nucleotide) in length between the left fusion gene and the right fusion gene in a window up to 4nt away from the breakpoint. It requires the homology to be anchored *at* the breakpoint.
### SSA
For single-strand annealing, FUSED looks for homology between 30 and 70nt in length in a window around the breakpoint up to 70nt away on both sides of the breakpoint.
### MMEJ
For microhomology-mediated end joining, FUSED looks for homology between 1-20nt in length between the left fusion gene and the right fusion gene in a window up to 50nt away from the breakpoint.
### BIR
For break-induced replication, FUSED looks for homology between 70-100nt in length in a window 100nt upstream and downstream of the breakpoint.
### MMBIR
For microhomology-mediated break-induced replication, FUSED looks for homology between 1-20nt in length in a window 20nt upstream and downstream of the breakpoint. It requires the homology to be anchored *at* the breakpoint, expects a gapped alignment of at least a single nt but with at least 80% overall homology.
## Input
FUSED expects a fusion prediction file from either: STAR, ARRIBA, or ORIEN derived pipelines. It utilizes a BSgenome object (or string) to pull matching sequence information from the coordinates of the predicted fusion. It can evaluate 5 different possible error repair mechanisms: NHEJ, MMEJ, SSA, BIR, or MMBIR.
## Usage
```R
devtools::install_github("databio/FUSED")
library(FUSED)
inputDT <- fread("ORIEN-FusionBreakpoints.txt")
calcHomology(inputDT, mechanism="NHEJ", bsg="BSgenome.Hsapiens.UCSC.hg38")
```