Error running convading: Incorrect BED file format, please check your BED file before processing

[lakhujanivijay]

I am getting 2 types of errors while running convading. My command is

perl/CoNVaDING-1.2.1/CoNVaDING.pl \
>   -mode StartWithBam \
>   -inputDir /00_bam_files \
>   -controlsDir 00_bam_files \
>   -outputDir 01_coverage_out \
>   -bed Target_annotated_intervals.bed

The errors are :

awk: cmd. line:1: fatal: division by zero attempted

Incorrect BED file format, please check your BED file before processing.

for fatal: division by zero, I know it is a known issue and I am waiting for the response here #28 (comment)

However, I don't know the reason about this error:

Incorrect BED file format, please check your BED file before processing.

Can you please help ?

[ljohansson]

Did you create your bed file according to the specifications? https://molgenis.gitbooks.io/convading/#create-normalized-count-files

Can you paste a few lines of your bedfile here?

[lakhujanivijay]

This is how my bed file looks

1	69090	70008	OR4F5
1	621095	622034	OR4F16
1	861321	861393	SAMD11

[mmterpstra]

It tries to match $line =~ m/.+\t[0-9]{1,}\t[0-9]{1,}\t[A-Za-z0-9]{1,}.+/gs
can you run the code below and paste the results here?

perl -wne 'warn "Invalid line !!!${_}!!! at $. " if(not(m/.+\t[0-9]{1,}\t[0-9]{1,}\t[A-Za-z0-9]{1,}.+/gs));' Target_annotated_intervals.bed

[lakhujanivijay]

@mmterpstra

Interestingly, the above command returns nothing.

[mmterpstra]

The code is confusing maybe run a dos2unix or try (this is more like the actual parsing of the file):

perl -wne 'my @bed =<ARGV>;my $i = 0; map {s/(?>\x0D\x0A?|[\x0A-\x0C\x85\x{2028}\x{2029}])//;chomp ;warn "Invalid line !!!${_}!!! at $i " if(not(m/.+\t[0-9]{1,}\t[0-9]{1,}\t[A-Za-z0-9]{1,}.+/gs)); $i++}(@bed);' Target_annotated_intervals.bed

If you got some results please post them it might help other people debugging their results.

[lakhujanivijay]

This time I got some errors:

perl -wne 'my @bed =<ARGV>;my $i = 0; map {s/(?>\x0D\x0A?|[\x0A-\x0C\x85\x{2028}\x{2029}])//;chomp ;warn "Invalid line !!!${_}!!! at $i " if(not(m/.+\t[0-9]{1,}\t[0-9]{1,}\t[A-Za-z0-9]{1,}.+/gs)); $i++}(@bed);' Target_annotated_intervals.bed
Invalid line !!!6	166571802	166572076	T!!! at 71623  at -e line 1, <> line 193137.
Invalid line !!!6	166574324	166574454	T!!! at 71624  at -e line 1, <> line 193137.
Invalid line !!!6	166575934	166576108	T!!! at 71625  at -e line 1, <> line 193137.
Invalid line !!!6	166578092	166578154	T!!! at 71626  at -e line 1, <> line 193137.
Invalid line !!!6	166578287	166578349	T!!! at 71627  at -e line 1, <> line 193137.
Invalid line !!!6	166579193	166579328	T!!! at 71628  at -e line 1, <> line 193137.
Invalid line !!!6	166580079	166580344	T!!! at 71629  at -e line 1, <> line 193137.
Invalid line !!!6	166580873	166581079	T!!! at 71630  at -e line 1, <> line 193137.

What does that mean?

[mmterpstra]

Yes! This helps:
the last '.+' isn't matched if you change the 'T' to a two letter combination then it will work.
lazy fix:

perl -wlape 'if(length($F[3]) == 1 ){$F[3].="_".$F[3];$_=join("\t",@F); warn "Fixed line number $. new content !!!$_!!!";}' old.bed > new.bed

[lakhujanivijay]

Sorry, I did not get you. How to I edit the BED file such that it works. What is wrong with the BED. Can you please elaborate ?

[mmterpstra]

The single letter name gives errors in convading. Also update the code above cause the gene name T should be changed to a constant name for the gene based normalsation. It now changes to T_T instead of T_$linenumber