-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Added shell script for generating the download page
- Loading branch information
Husen M. Umer
committed
Jul 5, 2018
1 parent
200b21f
commit 3128294
Showing
2 changed files
with
101 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,28 @@ | ||
|
||
<h4><p><b>Potential functional motifs per tissue type</b></p></h4><hr/><ul> | ||
<li><a href='bed/blood.funmotifs.bed.tar.gz' download>blood.funmotifs</a><br/></li> | ||
<li><a href='bed/brain.funmotifs.bed.tar.gz' download>brain.funmotifs</a><br/></li> | ||
<li><a href='bed/breast.funmotifs.bed.tar.gz' download>breast.funmotifs</a><br/></li> | ||
<li><a href='bed/cervix.funmotifs.bed.tar.gz' download>cervix.funmotifs</a><br/></li> | ||
<li><a href='bed/colon.funmotifs.bed.tar.gz' download>colon.funmotifs</a><br/></li> | ||
<li><a href='bed/esophagus.funmotifs.bed.tar.gz' download>esophagus.funmotifs</a><br/></li> | ||
<li><a href='bed/kidney.funmotifs.bed.tar.gz' download>kidney.funmotifs</a><br/></li> | ||
<li><a href='bed/liver.funmotifs.bed.tar.gz' download>liver.funmotifs</a><br/></li> | ||
<li><a href='bed/lung.funmotifs.bed.tar.gz' download>lung.funmotifs</a><br/></li> | ||
<li><a href='bed/myeloid.funmotifs.bed.tar.gz' download>myeloid.funmotifs</a><br/></li> | ||
<li><a href='bed/pancreas.funmotifs.bed.tar.gz' download>pancreas.funmotifs</a><br/></li> | ||
<li><a href='bed/prostate.funmotifs.bed.tar.gz' download>prostate.funmotifs</a><br/></li> | ||
<li><a href='bed/skin.funmotifs.bed.tar.gz' download>skin.funmotifs</a><br/></li> | ||
<li><a href='bed/stomach.funmotifs.bed.tar.gz' download>stomach.funmotifs</a><br/></li> | ||
<li><a href='bed/uterus.funmotifs.bed.tar.gz' download>uterus.funmotifs</a><br/></li> | ||
</ul> | ||
<hr/><h4><p><b>Annotated variatns using funMotifsDB</b></p></h4><hr/> | ||
<ul> | ||
<li>Putative regulatory variants from the 1000 Genomes project: <a href='../1000GenomeProject/functional/candidates1000G_tfexpr_entropyabs0.3_dnase_tfbindOrFscore2.55_grouped.tar.gz' download>candidate_variants_1000G.tar.gz</a>.</li> | ||
<li>Putative regulatory variants from the GTEx project (eQTLs): <a href='../GTEx_eQTLs/GTEx_eQTLs_candidates.tar.gz' download>GTEx_eQTLs_candidates.tar.gz</a>.</li> | ||
<li>Putative regulatory variants from GWAS : <a href='../GWAS_data/GWAS_SNPs_LD_candiadtes.tar.gz' download>GWAS_SNPs_LD_candiadtes.tar.gz</a>.</li> | ||
|
||
</ul> | ||
<br/><hr/><h4><p><b>All scored motifs per tissue type</b></p></h4><hr/> | ||
<ul><li><a href='bed/all_tissues.bed.tar.gz' download>All motifs: fscore per tissue type</a></li></ul><br/> | ||
<hr/><h4><p><b>Data files used to build funMotifsDB</b></p></h4><hr/><ul><li><a href='../datafiles.tar.gz' download>datafiles.tar.gz</a>: information about the contents of this archive is listed <a href='https://github.com/husensofteng/funMotifs/tree/master/ReadMe'>here</a>.</li></ul> |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,73 @@ | ||
downloadhtmlfile="../../download.html" | ||
db_name="funmotifsdb" | ||
tables_db_filename="table_names_funmotifsdb.txt" | ||
output_dir="www/" | ||
output_dir2="bed/" | ||
motifs_table="motifs" | ||
all_tissues_table="all_tissues" | ||
col_names_motifs="chr,motifstart,motifend,name,score,pval,strand" | ||
col_names_all_tissues="blood,brain,breast,cervix,colon,esophagus,kidney,liver,lung,myeloid,pancreas,prostate,skin,stomach,uterus" | ||
col_names_tissue_table="fscore, chromhmm, contactingdomain, dnase__seq, fantom, loopdomain, numothertfbinding, othertfbinding, replidomain, tfbinding, tfexpr" | ||
limit_stmt="" | ||
|
||
cd $output_dir$output_dir2 | ||
#get tissue names from the database | ||
psql -d $db_name -c "SELECT distinct(table_name) as table_name FROM information_schema.tables where table_schema='public' and table_name not like '%motif%' order by table_name;" | awk '$0!~"table" && $0!~"---" && $0!~"row" && $0!="" && $0!~"all_tissues"'> $tables_db_filename | ||
|
||
echo "" > $downloadhtmlfile | ||
#GENERATE filtered output (potential functional motifs) | ||
echo "<h4><p><b>Potential functional motifs per tissue type</b></p></h4><hr/><ul>" >> $downloadhtmlfile | ||
#export a file for each tissue table | ||
while read -r name | ||
do | ||
echo $output_dir$name.funmotifs | ||
psql -d $db_name -A -F $'\t' --pset footer -c "select $col_names_motifs,$col_names_tissue_table from $motifs_table,$name where $motifs_table.mid=$name.mid and dnase__seq>0 and (tfexpr>0 or tfexpr='nan') and (fscore>2.55 or (tfbinding>0 and tfbinding!='NaN'))$limit_stmt;" -o $name.funmotifs.bed | ||
tar -czvf $name.funmotifs.bed.tar.gz $name.funmotifs.bed | ||
echo "<li><a href='$output_dir2$name.funmotifs.bed.tar.gz' download>$name.funmotifs</a><br/></li>" >> $downloadhtmlfile | ||
done < $tables_db_filename | ||
echo "</ul>" >> $downloadhtmlfile | ||
|
||
|
||
|
||
#get tissue names | ||
tissues=$( | ||
while read line | ||
do | ||
echo -n ",${line}" | ||
done < $tables_db_filename) | ||
|
||
########variants###### | ||
echo "<hr/><h4><p><b>Annotated variatns using funMotifsDB</b></p></h4><hr/> | ||
<ul> | ||
<li>Putative regulatory variants from the 1000 Genomes project: <a href='../1000GenomeProject/functional/candidates1000G_tfexpr_entropyabs0.3_dnase_tfbindOrFscore2.55_grouped.tar.gz' download>candidate_variants_1000G.tar.gz</a>.</li> | ||
<li>Putative regulatory variants from the GTEx project (eQTLs): <a href='../GTEx_eQTLs/GTEx_eQTLs_candidates.tar.gz' download>GTEx_eQTLs_candidates.tar.gz</a>.</li> | ||
<li>Putative regulatory variants from GWAS : <a href='../GWAS_data/GWAS_SNPs_LD_candiadtes.tar.gz' download>GWAS_SNPs_LD_candiadtes.tar.gz</a>.</li> | ||
</ul>" >> $downloadhtmlfile | ||
|
||
##########get motif tables | ||
tables_motifs_db_filename="motifs_tables.txt" | ||
psql -d $db_name --pset footer -c "SELECT distinct(table_name) as table_name FROM information_schema.tables where table_schema='public' and table_name like '%motif%' and table_name like '%chr%' order by table_name;" | awk '$0~"chr" && $0~"motifs"' > $tables_motifs_db_filename | ||
|
||
echo "<br/><hr/><h4><p><b>All scored motifs per tissue type</b></p></h4><hr/>" >> $downloadhtmlfile | ||
|
||
#write the column headers to an outputfile | ||
psql -d $db_name -A -F $'\t' --pset footer -c "select $col_names_motifs$tissues from $motifs_table,$all_tissues_table where $motifs_table.mid=$all_tissues_table.mid limit 0;" > $all_tissues_table.bed | ||
|
||
while read -r name | ||
do | ||
echo $output_dir$name | ||
psql -d $db_name -A -F $'\t' -t -c "select $col_names_motifs$tissues from $name,$all_tissues_table where $name.mid=$all_tissues_table.mid$limit_stmt;" >> $all_tissues_table.bed | ||
done < $tables_motifs_db_filename | ||
tar -czvf $all_tissues_table.bed.tar.gz $all_tissues_table.bed | ||
|
||
|
||
echo "<ul><li><a href='$output_dir2$all_tissues_table.bed.tar.gz' download>All motifs: fscore per tissue type</a></li></ul><br/>" >> $downloadhtmlfile | ||
|
||
|
||
|
||
#####data_files.tar.gz | ||
echo "<hr/><h4><p><b>Data files used to build funMotifsDB</b></p></h4><hr/><ul><li><a href='../datafiles.tar.gz' download>datafiles.tar.gz</a>: information about the contents of this archive is listed <a href='https://github.com/husensofteng/funMotifs/tree/master/ReadMe'>here</a>.</li></ul>" >> $downloadhtmlfile | ||
|
||
|
||
|