split_file.R vs psiPerEvent number of fields expectation

[antpiron]

Dear Devs,

split_file.R expect a file formatted as (Note the column Name, this is generated with salmon quantmerge):

Name    hum54-1 hum54-2 hum54-3 hum54-4 hum56-2 hum56-3 hum56-4 hum57-1 hum57-2 hum57-3 hum57-4 mbe4-2  mbe4-3  mbe4-4
ENST00000640864       0       0       0       0       0       0       0       0       0       0       0       0       0       0
ENST00000638534       0       0       0       0       0       0       0       0       0       0       0       0       0       0

and psiPerEvent expects without the Name column name:

hum54-1 hum54-2 hum54-3 hum54-4 hum56-2 hum56-3 hum56-4 hum57-1 hum57-2 hum57-3 hum57-4 mbe4-2  mbe4-3  mbe4-4
ENST00000640864 0       0       0       0       0       0       0       0       0       0       0       0       0       0
ENST00000638534 0       0       0       0       0       0       0       0       0       0       0       0       0       0

With Name, I get:

ERROR:suppa.tools:35426, in line 35427. Unexpeced number of fields. 16 expected, 15 given. Skipping line...

Thanks,
Anthony.

[EduEyras]

Thanks, we should have fixed this a while back. JC, JL, what are the chances of making this work with the first column header? Thanks E.

…

On Mon, 25 Feb 2019 at 16:48, antpiron ***@***.***> wrote: split_file.R expect a file formatted as (Note the column Name): Name hum54-1 hum54-2 hum54-3 hum54-4 hum56-2 hum56-3 hum56-4 hum57-1 hum57-2 hum57-3 hum57-4 mbe4-2 mbe4-3 mbe4-4 ENST00000640864 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ENST00000638534 0 0 0 0 0 0 0 0 0 0 0 0 0 0 and psiPerEvent expects without the Name column name: hum54-1 hum54-2 hum54-3 hum54-4 hum56-2 hum56-3 hum56-4 hum57-1 hum57-2 hum57-3 hum57-4 mbe4-2 mbe4-3 mbe4-4 ENST00000640864 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ENST00000638534 0 0 0 0 0 0 0 0 0 0 0 0 0 0 With Name, I get: ERROR:suppa.tools:35426, in line 35427. Unexpeced number of fields. 16 expected, 15 given. Skipping line... — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#49>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AMWVB1XfjTkdwLxYsvtdpXmMwLiDpBPwks5vRAXNgaJpZM4bQLFa> .

-- Dr E Eyras ICREA Research Professor PRBB, Dr Aiguader 88 Tel: +34 93 316 0502 (ext 1502) E08003 Barcelona - Spain Fax: +34 93 316 0550 http://scholar.google.com/citations?user=LiojlGoAAAAJ http://www.researcherid.com/rid/L-1053-2014

[antpiron]

I think the problem is around

SUPPA/lib/tools.py

Lines 376 to 382 in 590060e

	fields = line.rstrip("\n").split("\t")
	if lineNumber == 0 and header: #Skip the header line
	#Calculating the number of fields required
	min_fields = (len(line.rstrip("\n").split("\t")) + 1)
	#Storing column_id for the expression fields
	colIds = fields
	continue

Should be something like

 fields = line.rstrip("\n").split("\t")
 if lineNumber == 0 and header: #Skip the header line
      #Calculating the number of fields required
      min_fields = len(fields)
      #Storing column_id for the expression fields
      colIds = fields[1:]

EDIT: I tested this code and it works. But to be coherent,

SUPPA/psiCalculator.py

Line 173 in 590060e

writer.writeLine("\t".join(col_ids), False)

should be

  writer.writeLine("\t".join(["Name"] + col_ids), False)

[antpiron]

This code is included in my pull request:
#48 (comment)

PS: careful, the pr change a lot of other things.

[antpiron]

My bad, the issue was not the name column but the minus - in the column names. check.names=FALSE in split_file.R solves the issue:

input_file <- read.table(CHARACTER_command_args[1],row.names=1,header=TRUE, check.names=FALSE)

Anyway, it will be nice to support the format outputted by salmon quantmerge.