Merge pull request #92 from leexgh/mutation-assessor-v4

Mutation assessor v4

data/grch37_ensembl92/export/annotation_version.txt

@@ -4,6 +4,6 @@ HGNC	2023-10	mirrored	hgnc	The resource for approved human gene nomenclature. Ge
 Cancer Hotspots	v2	mirrored	cancer_hotspots	A resource for statistically significant mutations in cancer 	https://www.cancerhotspots.org
 3D Hotspots	v2	mirrored	3d_hotspots	A resource for statistically significant mutations clustering in 3d protein structures in cancer	https://www.3dhotspots.org/
 reVUE	https://github.com/knowledgesystems/reVUE-data/blob/main/VUEs.json	mirrored	revue	A Repository for Variants with Unexpected Effects (VUE) in Cancer	https://www.cancerrevue.org/
-Mutation Assessor	v3	mirrored	mutation_assessor	Mutation Assessor predicts the functional impact of amino-acid substitutions in proteins, such as mutations discovered in cancer or missense polymorphisms.	http://mutationassessor.org/r3/
+Mutation Assessor	v4	mirrored	mutation_assessor	Mutation Assessor predicts the functional impact of amino-acid substitutions in proteins, such as mutations discovered in cancer or missense polymorphisms.	http://mutationassessor.org/r3/
 My Variant Info	Includes many annotation sources, see https://docs.myvariant.info/en/latest/doc/data.html	external	my_variant_info	MyVariant.info provides simple-to-use REST web services to query/retrieve variant annotation data, aggregated from many popular data resources.	https://myvariant.info
 ClinVar	20230722	mirrored	clinvar	ClinVar aggregates information about genomic variation and its relationship to human health.	https://www.ncbi.nlm.nih.gov/clinvar/

data/grch38_ensembl92/input/annotation_version.txt

@@ -0,0 +1,8 @@
+name	version	type	id	description	url
+VEP	grch38	mirrored	vep	VEP determines the effect of your variants(SNPs, insertions, deletions, CNVs or structural variants) on genes, transcripts, and protein sequence, as well as regulatory regions.	https://useast.ensembl.org/info/docs/tools/vep/index.html
+HGNC	2023-10	mirrored	hgnc	The resource for approved human gene nomenclature. Genome Nexus uses HGNC gene symbols in annotation	http://ftp.ebi.ac.uk/pub/databases/genenames/hgnc/archive/monthly/tsv/
+Cancer Hotspots	v2	mirrored	cancer_hotspots	A resource for statistically significant mutations in cancer 	https://www.cancerhotspots.org
+3D Hotspots	v2	mirrored	3d_hotspots	A resource for statistically significant mutations clustering in 3d protein structures in cancer	https://www.3dhotspots.org/
+Mutation Assessor	v4	mirrored	mutation_assessor	Mutation Assessor predicts the functional impact of amino-acid substitutions in proteins, such as mutations discovered in cancer or missense polymorphisms.	http://mutationassessor.org/r3/
+My Variant Info	Includes many annotation sources, see https://docs.myvariant.info/en/latest/doc/data.html	external	my_variant_info	MyVariant.info provides simple-to-use REST web services to query/retrieve variant annotation data, aggregated from many popular data resources.	https://myvariant.info
+ClinVar	20230722	mirrored	clinvar	ClinVar aggregates information about genomic variation and its relationship to human health.	https://www.ncbi.nlm.nih.gov/clinvar/

data/grch38_ensembl95/export/annotation_version.txt

@@ -3,5 +3,6 @@ VEP	grch38	mirrored	vep	VEP determines the effect of your variants(SNPs, inserti
 HGNC	2023-10	mirrored	hgnc	The resource for approved human gene nomenclature. Genome Nexus uses HGNC gene symbols in annotation	http://ftp.ebi.ac.uk/pub/databases/genenames/hgnc/archive/monthly/tsv/
 Cancer Hotspots	v2	mirrored	cancer_hotspots	A resource for statistically significant mutations in cancer 	https://www.cancerhotspots.org
 3D Hotspots	v2	mirrored	3d_hotspots	A resource for statistically significant mutations clustering in 3d protein structures in cancer	https://www.3dhotspots.org/
+Mutation Assessor	v4	mirrored	mutation_assessor	Mutation Assessor predicts the functional impact of amino-acid substitutions in proteins, such as mutations discovered in cancer or missense polymorphisms.	http://mutationassessor.org/r3/
 My Variant Info	Includes many annotation sources, see https://docs.myvariant.info/en/latest/doc/data.html	external	my_variant_info	MyVariant.info provides simple-to-use REST web services to query/retrieve variant annotation data, aggregated from many popular data resources.	https://myvariant.info
 ClinVar	20230722	mirrored	clinvar	ClinVar aggregates information about genomic variation and its relationship to human health.	https://www.ncbi.nlm.nih.gov/clinvar/

scripts/import_mongo.sh

@@ -52,22 +52,24 @@ elif [[ ${REF_ENSEMBL_VERSION} == *"grch38"* ]]; then
 fi
 
 # import mutation assessor
-if [[ ${REF_ENSEMBL_VERSION} == *"grch37"* && ${MUTATIONASSESSOR} == true ]]; then
-    echo "Downloading Mutation assessor data"
-
-    curl http://mutationassessor.org/r3/MA_scores_rel3_hg19_full.tar.gz -o ${DIR}/../data/common_input/MA_scores_rel3_hg19_full.tar.gz
-
-    echo "Extracting Mutation assessor data"
-    tar -xvf ${DIR}/../data/common_input/MA_scores_rel3_hg19_full.tar.gz
-    rm ${DIR}/../data/common_input/MA_scores_rel3_hg19_full.tar.gz
-
-    echo "Transforming Mutation assessor data"
-    sed -i -e 's/"Mutation","RefGenome variant","Gene","Uniprot","Info","Uniprot variant","Func. Impact","FI score"/_id,rgaa,gene,uprot,info,var,F_impact,F_score/g' MA_scores_rel3_hg19_full/MA_scores_rel3_hg19_chr*
-    sed -i -e 's/hg19,//g' MA_scores_rel3_hg19_full/MA_scores_rel3_hg19_chr*
-
-    echo "Importing Mutation assessor data"
-    for filename in MA_scores_rel3_hg19_full/*.csv; do import mutation_assessor.annotation $filename '--type csv --headerline' && rm $filename; done
-fi
+echo "Downloading Mutation assessor data"
+# Data source: https://drive.google.com/file/d/1V6r65xJFF5fJ7b9JHwqkvCe8wWDrIBhd/view. 
+# The copy is stored in S3 bucket: ttps://genome-nexus-static-data.s3.amazonaws.com/mutationassessor4_for_genome_nexus.tsv.xz
+curl https://genome-nexus-static-data.s3.amazonaws.com/mutationassessor4_for_genome_nexus.tsv.xz -o ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv.xz
+echo "Download completed."
+
+echo "Extracting Mutation assessor data"
+unxz -k ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv.xz
+
+echo "Transforming Mutation assessor data"
+sed -i '' 's/uniprotId\tSV\thgvspShort\tF_score\tF_impact\tMSA\tMAV/uniprotId\tsv\thgvspShort\tf_score\tf_impact\tmsa\tmav/' ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv
+awk -F'\t' 'BEGIN{OFS="\t"} NR==1{print "_id",$0; next} {print $1","$3,$0}' ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv > processed_mutaiton_assessor_tsv_file.tsv
+rm ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv.xz
+rm ${DIR}/../data/common_input/mutationassessor4_for_genome_nexus.tsv
+
+echo "Importing Mutation assessor data"
+import mutation_assessor.annotation processed_mutaiton_assessor_tsv_file.tsv "--type tsv --headerline"
+rm processed_mutaiton_assessor_tsv_file.tsv
 
 # import annotation sources version
 import version ${DIR}/../data/${REF_ENSEMBL_VERSION}/export/annotation_version.txt '--drop --type tsv --headerline'