problem with auto_annovar

[seqsyd]

I am trying to analyze some exome data using auto_annovar program. It works brilliantly till step 8. After Step-8, it generates a file but then it stop. During the step-8 processing, it says "fGrep: Writing output" and at the end the final message is "fGrep : Write error". Step-8 file contains the variants but with no gene names.
i am not able to map the variants to the relevant genes. I checked my Humandb database, it seems to contain all the relevant files. Can anyone please kindly help me with this. Thanks a lot

[Jane M]

Hello,

I am experiencing an issue with auto_annovar script too.
I am using annovar for a few months. Until now, I used annotate_variation.pl script, with different parameters separately, -geneanno, -regionanno, -filter,...
Now, I would like to try auto_annovar script, but I am stuck at the first step:

perl auto_annovar.pl -model recessive $dir/$file humandb -build hg19 -step 1
NOTICE: the --ver1000g argument is set as '1000g2010nov' by default
Error: the required database file humandb/hg19_ALL.sites.2010_11.txt does not exist. Please download it via -downdb argument by annotate_variation.pl.

I cannot find what this hg19_ALL.sites.2010_11.txt database is... I check the script and it is not mentioned in it.
I try to donload it:

perl annotate_variation.pl -downdb hg19_ALL.sites.2010_11.txt humandb -build hg19
NOTICE: Downloading annotation database ftp://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/hg19_ALL.sites.2010_11.txt.txt.gz ... ^[[AFailed
WARNING: Some files cannot be downloaded, including ftp://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/hg19_ALL.sites.2010_11.txt.txt.gz

perl annotate_variation.pl -downdb ALL.sites.2010_11.txt humandb -build hg19
NOTICE: Downloading annotation database ftp://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/ALL.sites.2010_11.txt.txt.gz ... Failed
WARNING: Some files cannot be downloaded, including ftp://hgdownload.cse.ucsc.edu/goldenPath/hg19/database/ALL.sites.2010_11.txt.txt.gz

Does someone know what is this file and where I can find it?

Thanks,
Jane

ps: I am using the version of March 2012.

[Heisman]

Try to include "-webfrom annovar" when downloading. Also, try "1000g2010nov" instead of "hg19_ALL.sites.2010_11.txt".

[Jane M]

Thank you for your answer, your remark put me on the right way: I have downloaded 1000g2012apr and not 1000g2010nov, so I only needed to change the name inside the script.

I managed to run the three first steps, but I don't really understand the outputs. I get files concerning almost all the steps: step4, 7, 8, 9, I don't know why, maybe they are here by defalut. Moreover, I cannot open the file .genelist, which seems to be the final file

perl auto_annovar.pl -model recessive $dir/$file humandb -build hg19 -step 1-3
NOTICE: the --ver1000g argument is set as '1000g2012apr' by default

NOTICE: Running step 1 with system command <perl annotate_variation.pl -geneanno -buildver hg19 -dbtype refgene -outfile GAR/mutect_file.step1 GAR/mutect_file humandb>
NOTICE: Reading gene annotation from humandb/hg19_refGene.txt ... Done with 40514 transcripts (including 6759 without coding sequence annotation) for 23468 unique genes
NOTICE: Reading FASTA sequences from humandb/hg19_refGeneMrna.fa ... Done with 63 sequences
WARNING: A total of 271 sequences will be ignored due to lack of correct ORF annotation
NOTICE: Finished gene-based annotation on 60 genetic variants in GAR/mutect_file
NOTICE: Output files were written to GAR/mutect_file.step1.variant_function, GAR/mutect_file.step1.exonic_variant_function

NOTICE: Running step 2 with system command <perl annotate_variation.pl -regionanno -dbtype mce46way -buildver hg19 -outfile GAR/mutect_file.step2 GAR/mutect_file.step2.varlist humandb>
NOTICE: Reading annotation database humandb/hg19_phastConsElements46way.txt ... Done with 5163775 regions
NOTICE: Finished region-based annotation on 26 genetic variants in GAR/mutect_file.step2.varlist
NOTICE: Output files were written to GAR/mutect_file.step2.hg19_phastConsElements46way

NOTICE: Running step 3 with system command <perl annotate_variation.pl -regionanno -dbtype segdup -buildver hg19 -outfile GAR/mutect_file.step3 GAR/mutect_file.step3.varlist humandb>
NOTICE: Reading annotation database humandb/hg19_genomicSuperDups.txt ... Done with 51599 regions
NOTICE: Finished region-based annotation on 24 genetic variants in GAR/mutect_file.step3.varlist
NOTICE: Output files were written to GAR/mutect_file.step3.hg19_genomicSuperDups

NOTICE: Running step 8 with system command <fgrep -f GAR/mutect_file.step8.varlist GAR/mutect_file.step1.exonic_variant_function | cut -f 2- > GAR/mutect_file.step8;cut -f 3- GAR/mutect_file.step8 > GAR/mutect_file.step8.temp;fgrep -v -f GAR/mutect_file.step8.temp GAR/mutect_file.step8.varlist > GAR/mutect_file.step8.temp1;fgrep -f GAR/mutect_file.step8.temp1 GAR/mutect_file.step1.variant_function >> GAR/mutect_file.step8;>

NOTICE: a list of potentially important genes and the number of variants in them are written to GAR/mutect_file.genelist
NOTICE: Consider filter out the list of dispensable genes from the GAR/mutect_file.genelist file to identify the final candidate gene list.

What do you get and is the .genelist file the final result?

[Heisman]

I don't actually use the auto_annovar script, I just use the summarize.pl script and do my own filtering from that, so I'm not sure exactly what you should get.

[Jane M]

Ok, thanks. I will try summarize.pl script since I don't really want to follow all the steps of auto_annovar.pl.