snpEff eff with built genome database

[htetre]

Hello,

I am having trouble annotating my vcf file using a database that I built. It seems as though snpEff wants to go to sourceforge.net to retieve a built database. Can someone help me identify what my problem is?

Here is my code
java -Xmx4g -jar /pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/snpEff.jar eff -c /pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/snpEff.config Sorghum EMSmutboth.vcf > var.ann.vcf

Here is the error I continue to receive:
ERROR while connecting to http://downloads.sourceforge.net/pro..._3_Sorghum.zip
java.lang.RuntimeException: java.lang.RuntimeException: File not found on the server. Make sure the database name is correct.
at org.snpeff.util.Download.download(Download.java:178)
at org.snpeff.snpEffect.commandLine.SnpEffCmdDownload.downloadAndInstall(SnpEffCmdDownload.java:32)
at org.snpeff.snpEffect.commandLine.SnpEffCmdDownload.runDownloadGenome(SnpEffCmdDownload.java:86)
at org.snpeff.snpEffect.commandLine.SnpEffCmdDownload.run(SnpEffCmdDownload.java:72)
at org.snpeff.SnpEff.run(SnpEff.java:1221)
at org.snpeff.SnpEff.loadDb(SnpEff.java:515)
at org.snpeff.snpEffect.commandLine.SnpEffCmdEff.run(SnpEffCmdEff.java:998)
at org.snpeff.snpEffect.commandLine.SnpEffCmdEff.run(SnpEffCmdEff.java:981)
at org.snpeff.SnpEff.run(SnpEff.java:1183)
at org.snpeff.SnpEff.main(SnpEff.java:162)
Caused by: java.lang.RuntimeException: File not found on the server. Make sure the database name is correct.
at org.snpeff.util.Download.download(Download.java:127)
... 9 more
java.lang.RuntimeException: Genome download failed!
at org.snpeff.SnpEff.loadDb(SnpEff.java:516)
at org.snpeff.snpEffect.commandLine.SnpEffCmdEff.run(SnpEffCmdEff.java:998)
at org.snpeff.snpEffect.commandLine.SnpEffCmdEff.run(SnpEffCmdEff.java:981)
at org.snpeff.SnpEff.run(SnpEff.java:1183)
at org.snpeff.SnpEff.main(SnpEff.java:162)

I built a database that seemed to work, so I am not sure what information I am missing. When I built the database I used the following command:
java -jar location/to/snpEff.jar build -gffs -v Sorghum

Thank you for any help

[finswimmer]

Hello,

how does you config file looks like? I guess snpEff cannot find your database and tries to download it.

Be sure to link to your local database in the config file or use the -dataDir flag to specify the folder where the database is located.

fin swimmer

[htetre]

fin swimmer,

Thank you for your reply.

This is what I have in my config file:

#Sorghum bicolor genome, version 3.0.1
Sorghum.genome : Sorghum

And this is the error I am getting when running snpEff eff:
Reading configuration file '/home/pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/snpEff.config'. Genome: 'Sorghum'
00:00:00 Reading config file: /home/pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/snpEff.config
00:00:01 done
00:00:01 Reading database for genome version 'Sorghum' from file '/home/pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/./data/Sorghum/snpEffectPredictor.bin' (this might take a while)
00:00:01 Database not installed

It seems that the snpEffectPredictor.bin file is not created. Which leads me to believe my database wasn't built correctly. However when I build the database I do not receive and major errors of a failure to build database.

Are you familiar with this problem?
Thanks for replying to original thread, I hope you know what is causing this error,
htetre

[finswimmer]

Hello htetre,

snpeff tries to find your database in this path:

/home/pkgs/snpeff-4.3.1p-1/share/snpeff-4.3.1p-1/./data/Sorghum/snpEffectPredictor.bin

As you can see there is a '.' in the path. You can define a absolute path in the config file for data_dir or use -dataDir flag as i suggest in my first post to link to the correct directory.

fin swimmer

[htetre]

Thank you fin swimmer,

I have changed that now, thanks to you. Now it is definitely going to the correct path. However I have determined that my database was never built correctly in that I do not have this snpEffectPredictor.bin file. When I try to 'find' it anywhere on the cluster it is not present.

Have you heard of that problem? Executing the following:
java -jar snpEff.jar build -gff3 -v Sorghum
starts the program and it is finding my genes.gff file but it is not following through and creating a predictor file.

Thank you
htetre

[finswimmer]

Hello,

i've never build a new database, so I cannot help here. Do you really need to build your own database? snpEff seems to have a prebuild database for Sorghum.

fin swimmer

[htetre]

Well, up to now my vcf file is based on the most recent Sorghum genome assembly and that assembly is not part of snpeff. But with the difficulty I am having to build genome database I'm thinking of redoing with an older assembly version that is part of the snpeff available databases. I was trying not to but seems as though the problem I am having is not common or its unfamiliar because I havent been able to find information.

Thanks
Hannah

[htetre]

Hello,

I have found my problem, and it resides in the gff file I have been using. It only contained the gene not gene_exon information. Once I used the correct gff file the genome database was built.