ePat (extended PROVEAN annotation tool) is a software tool that extends the functionality of PROVEAN: a software tool for predicting whether amino acid substitutions and indels will affect the biological function of proteins.

ePat extends the conventional PROVEAN to enable the following two things.

1. To calculate the pathogenicity of indel variants with frameshift and variants near splice junctions, for which the conventional PROVEAN could not calculate the pathogenicity of these variants.
2. Batch processing is used to calculate the pathogenicity of multiple variants in a mutation list (VCF file) in a single step.

In order to identify variants that are predicted to be functionally important from the mutation list, ePat can help filter out variants that affect biological functions by utilizing not only missense variants and indel variants that does not cause frameshift, but also frameshift, stop gain and splice variants.

• Ubuntu：Ubuntu20.04
• CentOS：CentOS7

• 10024MB or above

• Singularity 3.3.0 or above

Download from Zenodo

wget https://zenodo.org/record/5800418/files/ePat.zip

then unzip.

unzip ePat.zip

1. Create a working directory (YOUR_WORKDIR) and place a VCF file for input (YOUR_INPUTFILE), a FASTA file for reference genome (YOUR_REF_GENOME), and a GTF file for annotation (YOUR_REF_ANNO) in YOUR_WORKDIR. (HG38 is given as the default reference.) For example, if you want to annotate SARS-CoV-2 mutations, download a FASTA file and a GTF file from https://covid-19.ensembl.org/Sars_cov_2/Info/Index, then ungzip them.

# These are demo data
wget http://suikou.fs.a.u-tokyo.ac.jp/data/ePat/test_data2/input.vcf
wget http://suikou.fs.a.u-tokyo.ac.jp/data/ePat/test_data2/Sars_cov_2.ASM985889v3.dna.toplevel.fa
wget http://suikou.fs.a.u-tokyo.ac.jp/data/ePat/test_data2/Sars_cov_2.ASM985889v3.101.gtf

2. Make a temp directory to generate the intermediate fiels in YOUR_WORKDIR.

mkdir tmp

3. Execute the following command. (If you are searching for a gene for the first time, it will take a very long time to perform a BLAST search in the NCBI NR database.)

docker run -i --rm -v $PWD:$PWD -v $PWD/tmp:/root/tmp -w $PWD c2997108/epat:2 /usr/local/ePat/script/automated_provean.sh -i input.vcf -f Sars_cov_2.ASM985889v3.dna.toplevel.fa -g Sars_cov_2.ASM985889v3.101.gtf

4. After the analysis is finished, (YOUR_INPUTFILE)_dir/output/output_provean_(PREFIX_OF_YOUR_INPUTFILE).txt will be output as the output file. (Example: http://suikou.fs.a.u-tokyo.ac.jp/data/ePat/test_data2/output_provean_input.txt )
5. The 'PROVEAN_score' column shows the effect of the mutation on the protein function, and the 'PROVEAN_pred' column shows whether the mutation is harmful or not.

Download from Zenodo and unzip. (Use ePat/test_data directry as YOUR_WORKDIR)

wget https://zenodo.org/record/5482094/files/ePat.zip 

unzip ePat.zip

Make YOUR_TMPDIR

mkdir ePat/tmp

Check current directory (Use this output as PATH_TO_EPAT)

export PATH_TO_EPAT=$PWD 

Move to YOUR_WORKDIR

cd ePat/test_data

Run ePat

singularity run -B $PATH_TO_EPAT/ePat/test_data:$PATH_TO_EPAT/ePat/test_data -B $PATH_TO_EPAT/ePat/tmp:/root/tmp -W $PATH_TO_EPAT/ePat/test_data $PATH_TO_EPAT/ePat/ePat.sif /usr/local/ePat/script/automated_provean.sh -i input.vcf -f tmp.fa -g genes.gtf

If you want to use docker,

docker run -i --rm -v $WORKDIR:$WORKDIR -v $TMPDIR:/root/tmp -w $WORKDIR c2997108/epat:2 /usr/local/ePat/script/automated_provean.sh -i (YOUR_INPUTFILE) -f (YOUR_REF_GENOME) -g (YOUR_REF_ANNO)

Check Result

cat $PATH_TO_EPAT/ePat/test_data/input.vcf_dir/output/output_provean_input.txt

#CHROM  POS       ID           REF  ALT  QUAL   FILTER  INFO                                                                                                                                                                                                                                                                  FORMAT  HG00096  PROVEAN_pred  PROVEAN_score
22      24698294  rs202142165  C    G    100.0  PASS    AC=1;AF=0.000199681;AN=5008;NS=2504;DP=18536;EAS_AF=0;AMR_AF=0;AFR_AF=0.0008;EUR_AF=0;SAS_AF=0;AA=C|||;VT=SNP;EX_TARGET;ANN=G|missense_variant|MODERATE|SPECC1L|SPECC1L|transcript|NM_015330|protein_coding|3/17|c.95C>G|p.Ser32Cys|389/6763|95/3354|32/1117||        GT      0|0      N             -1.005
22      24709317  rs548017612  G    A    100.0  PASS    AC=1;AF=0.000199681;AN=5008;NS=2504;DP=20098;EAS_AF=0;AMR_AF=0;AFR_AF=0;EUR_AF=0;SAS_AF=0.001;AA=G|||;VT=SNP;EX_TARGET;ANN=A|missense_variant|MODERATE|SPECC1L|SPECC1L|transcript|NM_015330|protein_coding|4/17|c.190G>A|p.Gly64Arg|484/6763|190/3354|64/1117||       GT      0|0      N             -2.165
22      24709321  rs569594248  G    C    100.0  PASS    AC=1;AF=0.000199681;AN=5008;NS=2504;DP=19830;EAS_AF=0;AMR_AF=0;AFR_AF=0;EUR_AF=0.001;SAS_AF=0;AA=G|||;VT=SNP;EX_TARGET;ANN=C|missense_variant|MODERATE|SPECC1L|SPECC1L|transcript|NM_015330|protein_coding|4/17|c.194G>C|p.Gly65Ala|488/6763|194/3354|65/1117||       GT      0|0      N             -0.881
22      24709420  rs35783914   C    T    100.0  PASS    AC=4;AF=0.000798722;AN=5008;NS=2504;DP=19262;EAS_AF=0;AMR_AF=0.0014;AFR_AF=0;EUR_AF=0.003;SAS_AF=0;AA=C|||;VT=SNP;EX_TARGET;ANN=T|missense_variant|MODERATE|SPECC1L|SPECC1L|transcript|NM_015330|protein_coding|4/17|c.293C>T|p.Ser98Phe|587/6763|293/3354|98/1117||  GT      0|0      N             -1.051
22      24709423  rs371780453  A    G    100.0  PASS    AC=1;AF=0.000199681;AN=5008;NS=2504;DP=19367;EAS_AF=0;AMR_AF=0;AFR_AF=0;EUR_AF=0;SAS_AF=0.001;AA=A|||;VT=SNP;EX_TARGET;ANN=G|missense_variant|MODERATE|SPECC1L|SPECC1L|transcript|NM_015330|protein_coding|4/17|c.296A>G|p.Lys99Arg|590/6763|296/3354|99/1117||       GT      0|0      N             -0.388

If you want to share a database built with snpEff or sss files aligned with provean to SHARED_DIR.　The following commands can be run after the above command using the options "-f" and "-g" has been executed.

### Change follows to fit your environment. ###
i=input.vcf
SHARED_DIR=$PATH_TO_EPAT/ePat/test_data/input.vcf_dir/snpEff
WORK_DIR=$PATH_TO_EPAT/ePat/test_data
TMP_DIR=/tmp/$i
###############################################
mkdir -p $TMP_DIR 
singularity run -B $TMP_DIR:/root/tmp -B $SHARED_DIR:/root/snpEff -B $WORK_DIR:$WORK_DIR -W $WORK_DIR $PATH_TO_EPAT/ePat/ePat.sif /usr/local/ePat/script/automated_provean.sh -i $i -r tmp
rm -rf $TMP_DIR

The input data is a VCF file after variant call, a FASTA file of the reference genome, and a GTF file with gene annotations.

With given reference, we create a database for SnpEff and annotate the VCF file with SnpEff. We then extract variants that have a HIGH or MODERATE pathogenicity level as a result of the SnpEff annotation.

For each row of the VCF file, extract the information of the variants annotated with SnpEff ([gene ID, variant type, pathogenicity level, DNA mutation information, amino acid mutation information]) from the INFO column. With this information, the variants are classified into (1) variants near the splice junction(splice variants), (2) frameshift, (3) Stop Gain, (4) Start Lost, and (5) inframe variants (point Mutation or indel mutations that do not cause frameshift).

Variants from (1) to (4) are given pathogenicity as defined by ePat, and those (5) will be given pathogenicity by PROVEAN. The pathogenicity defined by ePat is calculated with the following method.

For each position, calculate the pathogenicity when it is replaced by each of the 20 amino acids. The average of these pathogenicity is used as the pathogenicity for that position. The maximum pathogenicity for each position is the pathogenicity of this mutation.

Calculate the pathogenicity defined by ePat in the range from the splice junction where the mutation occurs to the stop codon.

variants annotated as sequence_feature (due to a bug in SnpEff that annotates the pathogenicity as HIGH) and variants occuring in introns after the stop codon are not given the pathogenicity.

Pathogenicity defined by ePat is calculated in the range from the amino acid where the frameshift starts to the stop codon.

Calculate the pathogenicity defined by ePat in the range from the amino acid to be replaced by the stop codon to the original stop codon.

For Stop Lost, the pathogenicity is not calculated.

Calculate the pathogenicity defined by ePat in the range from the original start codon to the next methionine.

Calculate the pathogenicity by PROVEAN.

Assign these scores to the PROVEAN_score column, and assign D (Damaged) if the score is less than -2.5, or N (Neutral) if the score is greater than -2.5 to the PROVEAN_pred column.

The output is output as output_provean_{PREFIX_OF_YOUR_INPUTFILE}.txt and saved in the output directory.