examples

Example files to for testing PrecisionProDB

Variant files

• a small vcf file: celline.vcf.gz
• a small tsv file: gnomAD.variant.txt.gz

For more larger files, users may find tsv files based on gnomAD 3.1 data from
https://github.com/ATPs/PrecisionProDB_references/tree/main/gnomAD3.1

A bigger VCF file from Jurkat cellline (including the one with T2T CHM13 as reference genome) can be found
https://github.com/ATPs/PrecisionProDB_references/tree/main/cellline

Jurkat.CHM13.RefSeq.vcf.gz

Run with only local files

Users can run PrecisionProDB with local genome files and gene models, which can is faster than the running mode of downloading the required files from the Internet.

Example files for each data types

• Ensembl
• GENCODE
• RefSeq
• TransDecoder
• UniProt

Check each of the subfolder to see how to run the testing codes.

Run with the variant file only

PrecisonProDB can download the latest version of required files automatically from the Internet, so users can run get the results with the variant file only. However, as downloading large files may take some time, the running time will be longer.

Check this page for explanation of output files
https://github.com/ATPs/PrecisionProDB/wiki/Outputs-of-PrecisionProDB

enable sqlite

cd $PATH_OF_PRECISONPRODB
cd examples
python ../src/PrecisionProDB.py -m gnomAD.variant.txt.gz -D Ensembl -o ./PREFIX.tsv -a Ensembl_GTF --PEFF \
                    -S path_to_file_sqlite

If path_to_file_sqlite does not exist, a sqlite file will be created.

If path_to_file_sqlite is provided, the program will use informations stored in the sqlite file to speed up.

file path_to_file_sqlite can be build in advance with the following command.

cd $PATH_OF_PRECISONPRODB
python ./src/buildSqlite.py  -g GRCh38.p14.genome.fa.gz \
                    -p gencode.v47.pc_translations.fa.gz \
                    -f gencode.v47.chr_patch_hapl_scaff.annotation.gtf.gz \
                    -o GENCODE.tsv.buildSqlite \
                    -a GENCODE_GTF \
                    -S GENCODE.sqlite \
                    -t 8

variants in tsv format

cd $PATH_OF_PRECISONPRODB
cd examples
python ../src/PrecisionProDB.py -m gnomAD.variant.txt.gz -D Ensembl -o ./PREFIX.tsv -a Ensembl_GTF --PEFF

Here --PEFF means the PEFF option is enabled, and a result protein in PEFF format with VariantSimple annotations will be generated.

-D specified the gene models to use and to be downloaded. It can be GENCODE, RefSeq, Ensembl or Uniprot.

To run with your own data, just change gnomAD.variant.txt.gz to the location of your variant file.

Output files

• PREFIX.tsv.pergeno.aa_mutations.csv
• PREFIX.tsv.pergeno.protein_all.fa
• PREFIX.tsv.pergeno.protein_changed.fa
• PREFIX.tsv.pergeno.protein_PEFF.fa

variants in vcf format

cd $PATH_OF_PRECISONPRODB
cd examples
python ../src/PrecisionProDB.py -m celline.vcf.gz -D Ensembl -o ./PREFIX.vcf -a Ensembl_GTF --PEFF

Output files:

• PREFIX.vcf.pergeno.aa_mutations.csv
• PREFIX.vcf.pergeno.protein_all.fa
• PREFIX.vcf.pergeno.protein_changed.fa
• PREFIX.vcf.pergeno.protein_PEFF.fa
• PREFIX.vcf.vcf2mutation_1.tsv
• PREFIX.vcf.vcf2mutation_2.tsv

Here, the -m is set to be celline.vcf.gz.

If the vcf provided by the user contails multiple samples, -s/--sample option can be used to specify a specific sample. If not set, the first sample in the vcf file will be used.

Uniprot

For the two examples above, if -D Uniprot is set, several more output files will be generated.

• Uniprot.uniprot_all.fa
• Uniprot.uniprot_changed.fa
• Uniprot.uniprot_changed.tsv
• Uniprot.uniprot_PEFF.fa

These files should be used for downstream analysis.

variant in string format

cd $PATH_OF_PRECISONPRODB

python ./src/PrecisionProDB.py \
        -m chr1-942451-T-C,1-6253878-C-T,1-2194700-C-G,1-1719406-G-A \
        -o ./test_output/Ensembl/str/sqlite_two_step/Ensembl.str.sqlite_two_step \
        -a Ensembl_GTF --PEFF -t 4 \
        -S ./test_output/Ensembl/Ensembl.sqlite

two output will be generated

• Ensembl.str.sqlite_two_step.pergeno.aa_mutations.csv
• Ensembl.str.sqlite_two_step.pergeno.mutated_protein.fa

Note: if variant is a string, the -S option is required.

CHM13 T2T from RefSeq

Since the data is in the same format like RefSeq, we tested the command below and it works perfectly.

cd $PATH_OF_PRECISONPRODB

python ./src/downloadHuman.py -d CHM13 -o precisionprodb

python ./src/PrecisionProDB.py \
        -g precisionprodb/GCF_009914755.1_T2T-CHM13v2.0_genomic.fna.gz \
        -p precisionprodb/GCF_009914755.1_T2T-CHM13v2.0_protein.faa.gz \
        -t 12 \
        -a RefSeq \
        -f precisionprodb/GCF_009914755.1_T2T-CHM13v2.0_genomic.gtf.gz \
        -m precisionprodb/Jurkat.CHM13.RefSeq.vcf.gz \
        -o precisionprodb/Jurkat.CHM13.RefSeq.PrecisonProDB \
        -S precisionprodb/GCF_009914755.1_T2T-CHM13v2.0.sqlite

vcf2mutation

to get tsv file from the vcf file, you can use the vcf2mutation.py script

cd $PATH_OF_PRECISONPRODB
python ./src/vcf2mutation.py -h

usage: vcf2mutation [-h] -i FILE_VCF [-o OUTPREFIX] [-s SAMPLE]
                    [-F] [-A]

convert extract mutation information from vcf file

options:
  -h, --help            show this help message and exit
  -i FILE_VCF, --file_vcf FILE_VCF
                        input vcf file. It can be a gzip file
  -o OUTPREFIX, --outprefix OUTPREFIX
                        output prefix to store the two output
                        dataframes, default: None, do not write
                        the result to files. file will be
                        outprefix_1/2.tsv
  -s SAMPLE, --sample SAMPLE
                        sample name in the vcf to extract the
                        variant information. default: None,
                        extract the first sample
  -F, --no_filter       default only keep variant with value
                        "PASS" FILTER column of vcf file. if
                        set, do not filter
  -A, --all_chromosomes
                        default keep variant in chromosomes and
                        ignore those in short fragments of the
                        genome. if set, use all chromosomes
                        including fragments