randomMut

RandomMut is a python package to randomize mutations within a defined window size and preserving the trinucleotide context of the mutation.

Install

Install from PyPi:

pip install randommut

or cloning this repo:

git clone URL
cd randommut
python setup.py install

If the install was succesful you should see the help message when executing:

randommut -h

Usage

The first step is to serialize your genome and store it as a python object. This will actually generate a bigger file but it does save time in the long run. The amount of time saved doing this step will depend in the IO speed. It is convininent because you can serialize your refseq genome and then use the same to randomize more data sets.

GENOME="path/to/refseq.fa"
python -m randommut -M serialize -g $GENOME -a hg19

It should output the serialized version of the genome in the current directory. In terms of memory, this processed has a peak of memory when writting the file approx. at 30Gb for the hg19 assembly.

The second step is to generate the random positions. We will input the mutations in table format (see below).

• The times argument (-t) is equivalent to the number of randomizations.
• The winlen argument (-w) is equivalent to the length of the windows where the positions will be generated.

GENOME="path/to/refseq.fa.p"
MUTS="path/to/muts.tsv"
OUTFILE="path/to/outfile.tsv"
python random_genome_classic.py -M randomize -g $GENOME -m $MUTS -a hg19 -o $OUTFILE -t 50 -w 50000

Preprocessing

The input format is defined as a table (tsv format) with the folloing columns:

• chr: chromosome name (should match the chromsoma name in the fasta file)
• start: position start (1-based)
• end: position end
• ref: reference allele
• alt: alternative allele
• strand: Not used currently
• sample: Not relevant as each mutation is independent

chr1    10  10  G   A   1   SAMPLE1
chr2    20  20  C   T   1   SAMPLE2
...

Convert to table format

• From ICGC data release files.

ICGC="path/to/ICGC"
zcat $ICGC | awk 'BEGIN {FS="\t";OFS="\t"} $14~/single/{$9="chr"$9;print($9,$10,$11,$16,$17,$12,$5);}'

• From a vcf (unisample) we can use bcftools

bcftools query -f '%CHROM\t%POS\t%POS\t%REF\t%ALT{0}\t1\tsampleA\n' file.vcf > "file.tsv"

Post-processing

Sometimes do you need to perform small actions to adecuate the output.

Remove NNN positions

awk '$8!~/N/ {print($0);}' output_file.tsv

Check results are in the winlen range

Open the file with Excel.

Open conditional formating, custom rule

=AND(H2>($B2-50000);H2<($C2+50000))

Change 50000 with your inputed winlen parameter.