Error building snpEff database "Transcript 'hypothetical_protein' already exists"

Question

[Lina F]

Hi all,

I am trying to build a snpEff database but I'm running into the following error message:

java.lang.RuntimeException: Transcript  'hypothetical_protein' already exists
        at org.snpeff.snpEffect.factory.SnpEffPredictorFactory.add(SnpEffPredictorFactory.java:135)
        at org.snpeff.snpEffect.factory.SnpEffPredictorFactoryFeatures.addMrna(SnpEffPredictorFactoryFeatures.java:183)
        at org.snpeff.snpEffect.factory.SnpEffPredictorFactoryFeatures.addFeatures(SnpEffPredictorFactoryFeatures.java:134)
        at org.snpeff.snpEffect.factory.SnpEffPredictorFactoryFeatures.create(SnpEffPredictorFactoryFeatures.java:330)
        at org.snpeff.snpEffect.commandLine.SnpEffCmdBuild.run(SnpEffCmdBuild.java:369)
        at org.snpeff.SnpEff.run(SnpEff.java:1183)
        at org.snpeff.SnpEff.main(SnpEff.java:162)
java.lang.RuntimeException: Error reading file '/home/lina/snpEff/./data/my_organism/genes.gbk'
java.lang.RuntimeException: Transcript  'hypothetical_protein' already exists
        at org.snpeff.snpEffect.factory.SnpEffPredictorFactoryFeatures.create(SnpEffPredictorFactoryFeatures.java:344)
        at org.snpeff.snpEffect.commandLine.SnpEffCmdBuild.run(SnpEffCmdBuild.java:369)
        at org.snpeff.SnpEff.run(SnpEff.java:1183)
        at org.snpeff.SnpEff.main(SnpEff.java:162)

The file I am using for the database is a Genbank file that I downloaded from NCBI. It contains 12760 genes and 2764 of them are annotated with product="hypothetical protein"

Based on the error message I assume having more than one gene labeled hypothetical protein is a problem for snpEff. However, I assume there must be many organisms where that is the case.

Does anyone have any insight into this?

Thanks!

~Lina

[Pierre Lindenbaum]

what the acn of this genbank file ?

[Lina F]

it's GCA_001007165.2

Answer 1

[Pierre Lindenbaum]

It's because the name of the gene 'hypothetical protein" should be uniq

you can try (not tested)

 awk  '($0 ~ /product="hypothetical protein"/) {gsub(/ein/,"ein"NR);print;} {print;}'  in.gb > fixex.gb

[Lina F]

Thank you, this worked! However, now it's finding other non-unique names... looks like I will have to run this fix a few times. Thanks!

[Jane]

Hi, I have the same issue. After running the command, I run the "sh scripts/buildDbNcbi.sh U00096.2" to download again. Then the same issue appears. May I ask for the details of your operation? Thank you very much! Jane