search

weberlab-hhu / Helixer

Using Deep Learning to predict gene annotations
GNU General Public License v3.0
140 stars 20 forks source link

Import into Geenuff can show too high masked rate #47

Closed soi closed 3 years ago

soi commented 3 years ago

The masked rate can go > 100%. Maybe the count is not reset on coordinates without annotations.

(sorry, my bad about the first bug report, general import seems fine)

soi commented 3 years ago

1 expected num of chunks to write in 10000000000 bases to hdf5 Numerification of only the sequence of scaffold_1034 took 0.00 secs WARNING:root:Sequence scaffold_1034 has no annotations WARNING:root:Sequence scaffold_1034 has no annotations 443/853 Numerified <Coordinate 21, seqid: scaffold_1034, len: 3835, genome_id: 1, genome: crubella> of crubella with 0 features in 1 chunks, masked rate: 421.51%, ig rate: 0.00%, filtered fully err chunks: 0.00% (test), filtered chunks from featureless coordinates 0.00% (0.01 secs)

soi commented 3 years ago

fixed