Hi!

I'm using SnpEff to annotate SNP from Varscan. Since I'm working on Casava, I have build a database locally using Casava genome and its annotation in GFF format. When inspecting the database snpEff reports:

snpEff.jar dump snpEff_db | less

...

#-----------------------------------------------
# Genome name                : 'snpEff_db'
# Genome version             : 'snpEff_db'
# Genome ID                  : 'snpEff_db[0]'
# Has protein coding info    : true
# Has Tr. Support Level info : true
# Genes                      : 80115
# Protein coding genes       : 72330
#-----------------------------------------------
# Transcripts                : 95959
# Avg. transcripts per gene  : 1.20
# TSL transcripts            : 0
#-----------------------------------------------
# Checked transcripts        : 
#               AA sequences :      0 ( 0.00% )
#              DNA sequences :      0 ( 0.00% )
#-----------------------------------------------
# Protein coding transcripts : 87317
#              Length errors :    545 ( 0.62% )
#  STOP codons in CDS errors :      0 ( 0.00% )
#         START codon errors :  87310 ( 99.99% )
#        STOP codon warnings :  86772 ( 99.38% )
#              UTR sequences :  40611 ( 42.32% )
#               Total Errors :  87310 ( 99.99% )
#-----------------------------------------------
# Cds                        : 275258
# Exons                      : 335821
# Exons with sequence        : 335821
# Exons without sequence     : 0
# Avg. exons per transcript  : 3.50
# WARNING                    : No mitochondrion chromosome found
#-----------------------------------------------

Almost all protein coding gene have no START, STOP or both! I have also tried with the GTF version of the annotation. Since some annotation are incomplete or corrupted, I have manually inspected CDS position using IGV, and have not seen errors. Does some already faced this issue? How did you fixed it?

Thanks for help,