User:Lindenb/Notebook/UMR915/20110225

Problem with "Variation predictor:" answer from the dev mailing list: see http://lists.ensembl.org/pipermail/dev/2011-February/000928.html

Submitting each SNP after each SNP :-( UGLY

rm -f merged_result.vcf cat  ${HOME}/sample1.hg19.vcf ${HOME}/sample2.hg19.vcf | cut -d '	' -f 1-5 | grep -v "#" | sort | uniq > jeter.vcf while read line do echo "$line" echo $line | ./variant_effect_predictor.pl --format vcf -o jeter.variant .txt --hgnc --check_ref cat jeter.variant.txt | grep -v "Uploaded Variation" >> merged_result.v cf	rm jeter.variant.txt done < jeter.vcf

Problem with the web interface of Ensembl:

SNPs missing in the variant effect predictor (WWW interface) - www.ensembl.org

Hello Pierre,

Basically the limit of 750 is meant to apply to the number of features uploaded. However, after some investigation it turns out that the limit in fact seems to be applied on the number of features being returned; since one variation can produce multiple features in the results, the results are getting truncated early (i.e. when the number of output lines is 750, not when the number of input lines reached is 750).

This is obviously not ideal and goes against what it says on the site, so we will try and get this patched ASAP.

In the meantime, the only solution is to chop the file into smaller chunks. I would suggest 100 per file to be sure (there should be ~2400 rows returned from this file, ~5x the number of input rows).

My apologies for any inconvenience caused.