Skip to content

Latest commit

 

History

History
59 lines (42 loc) · 3.09 KB

README_puhti.md

File metadata and controls

59 lines (42 loc) · 3.09 KB

muat

Mutation-Attention Model

How to use:

Preprocessing:

A) Prepare the environment:

conda env create -f muat-conda.yml

B) Download dataset:

  1. run scripts/download_icgc.sh to download the data
  2. download genome reference --> http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/phase2_reference_assembly_sequence/hs37d5.fa.gz

C) Preprocessing Steps:

  1. copy scripts/dmmpre.sh, scripts/preprocessdmm.sh, scripts/annotate_preprocessed.sh to <data_dir>/release_28/
  2. edit dmmpre.sh assigning to dependencies (check the dmmpre.sh for information)
  3. run preprocessdmm.sh > if successful, this will give an output of 'somatic_mutations.<icgc_project>.dmm.k256.tsv.gz' in every icgc project folder
  4. annotate_preprocessed.sh > ex: ./annotate_preprocessed.sh ./BLCA-US/somatic_mutations.BLCA-US.dmm.k256.tsv.gz BLCA-US if successful, this file will give an output of somatic_mutations.<icgc_project>.dmm.k256.annotated.tsv.gz in every icgc project folder

D) Create dataset for MuAt

  1. run preprocessing/create_PCAWG_dataset.py (adjust the dependencies accordingly)

DEMO I) Training from the scratch

python3 main.py --dataloader 'pcawg' --block-size 5000 --n-class 24 --n-layer 1 --n-head 1 --n-emb 256 --motif --mut-type 'SNV+MNV' --fold 1 --input-data-dir '/path/to/input/data/dir/' --save-ckpt-dir '/path/to/saveckptdir/' --train

II) Predicting from I

python3 main.py --dataloader 'pcawg' --block-size 5000 --context-length 3 --n-class 24 --n-layer 1 --n-head 1 --n-emb 256 --motif --mut-type 'SNV+MNV' --fold 1 --input-data-dir '/path/to/input/data/dir/' --load-ckpt-dir '/path/to/saveckptdir/' --predict

III) Predicting from pretrained model

python3 main.py --dataloader 'pcawg' --block-size 5000 --context-length 3 --n-class 24 --n-layer 2 --n-head 1 --n-emb 512 --motif-pos --mut-type 'SNV+MNV+indel' --fold 1 --input-data-dir '/path/to/input/data/dir/' --load-ckpt-dir './bestckpt/fullpcawgfold1_11100_wpos_TripletPosition_bs5000_nl2_nh1_ne512_cl3/' --predict

IV) Predicting vcf files from pretrained model

python3 main.py --dataloader 'pcawg' --input-data-dir '/path/to/muat/data/raw/vcf/' --input-filename '00b9d0e6-69dc-4345-bffd-ce32880c8eef.consensus.20160830.somatic.snv_mnv.vcf' --tmp-dir '/path/to/muat/data/raw/temp/' --reference '/path/to/genome_reference/files' --load-ckpt-dir '/path/to/muat/bestckpt/wgs/' --load-ckpt-filename 'motif+position_features.pthx' --output-pred-dir '/path/to/muat/data/raw/outputdir/' --output-prefix 'test' --single-pred-vcf --get-features

Notes: --get-features, --output-prefix are optional

V) Predict all vcf files in the directory from pretrained model

python3 main.py --dataloader 'pcawg' --input-data-dir '/path/to/muat/data/raw/vcf/' --tmp-dir '/path/to/tempdir/' --reference '/path/to/genome_reference/files' --load-ckpt-dir '/path/to/muat/bestckpt/wgs/' --load-ckpt-filename 'motif+position_features.pthx' --output-pred-dir '/path/to/muat/data/raw/outputdir/' --output-prefix 'test' --multi-pred-vcf --get-features

Notes: --get-features, --output-prefix are optional