I built a snpEff database for my nuclear genome. It worked. I am now trying to build own for the chloroplast and mitochondria genomes. I've edited the config file and I have fasta files in the genome folder and their corresponding .gff files in their respective folders. I'm getting this error after I run the build command:
Reading GFF3 data file : '/home/kmmahan/snpEff/./data/scenedesmus_obliquus_doe0152z_mitochondria/genes.gff'
java.lang.StackOverflowError
I checked the fasta file and the gff file and the genome names match. It's not a memory error because I was able to build the nuclear genome database. I get the error for both the chloroplast and mitochondria database. I've searched for answers on biostars but I am stuck. Something seems to be wrong when it tries to read the gff file. What changes can I make?
It's the same path I used for the nuclear genome (except folder was scenedesmus_obliquus_doe0152z_nuclear/genes.gff). And same way that it was done here: A: Building Snpeff Database
I cd's using the path you listed (/home/kmmahan/snpEff/./data/scenedesmus_obliquus_doe0152z_mitochondria) and was taken to that directory. I think the path is correct.
Oh, okay. What did you use to read the gff file? That program probably makes a lot of recursive calls, that fills up the stack.
What do you mean read? I just entered the build command to build the database
11. java -jar snpEff.jar build -gff3 -v scenedesmus_obliquus_doe0152z_mitochondria
Oh maybe its the -gff3? Except I used the same thing to build the database for the nuclear genome . . .
Yes, -gff3 probably calls a different set of functions.
If I am using a gff file- I have to put -gff3 in the build command. Also - I used -gffs to build the database for the nuclear genome and it worked. Something seems to be wrong with the gff files themselves but just haven't figured it out.