gemBS extract bed file has incorrect delimiters

[paul-sud]

In recent updates to the code it looks like the gemBS style bed files have an extra space inserted as a delimiter after the third column after the feature end coordinate, in addition to the usual tab. Oddly, this doesn't seem to occur with the ENCODE-style bed files. tabix also seemed to run successfully on the file too. The data here was produced by running a subset of the test set from here: http://statgen.cnag.cat/GEMBS/UserGuide/_build/html/example.html

ENCODE-style bed files:

> head -n 10 sample_sample5_cpg.bed | tr '  '  '#'
track#name="sample5"#description="sample5"#visibility=2#itemRgb="On"
chrI    144 145 "sample5"   0   +   144 145 0,255,0 0   0   CG  CG  80
chrI    145 146 "sample5"   27  -   145 146 0,255,0 27  0   CG  CG  2
chrI    175 176 "sample5"   0   +   175 176 0,255,0 0   0   CG  CG  107
chrI    176 177 "sample5"   39  -   176 177 0,255,0 39  0   CG  CG  2
chrI    216 217 "sample5"   0   +   216 217 0,255,0 0   0   CG  CG  107
chrI    217 218 "sample5"   36  -   217 218 0,255,0 36  0   CG  CG  2
chrI    325 326 "sample5"   0   +   325 326 0,255,0 0   0   CG  CG  80
chrI    326 327 "sample5"   28  -   326 327 0,255,0 28  0   CG  CG  2
chrI    339 340 "sample5"   1   +   339 340 0,255,0 1   0   CG  CG  73

gemBS-style bed file, note the # after the feature stop coordinate:

> head -n 10 sample_sample5_cpg.txt | tr  '  '  '#'
Contig  Pos0    Pos1    Ref sample_sample5:Call sample_sample5:Flags    sample_sample5:Meth sample_sample5:non_conv sample_sample5:conv sample_sample5:support_call sample_sample5:total
chrI    834 835#    C   C   GQ=72;MQ=60 0.000   0   4   39  40
chrI    835 836#    G   G   GQ=13;MQ=60 0.019   1   35  40  40
chrI    868 869#    C   C   GQ=81;MQ=60 0.000   0   4   31  31
chrI    869 870#    G   G   GQ=12;MQ=60 0.000   0   28  32  32
chrI    890 891#    C   C   GQ=87;MQ=60 0.000   0   4   33  33
chrI    891 892#    G   G   GQ=13;MQ=60 0.026   1   28  33  33
chrI    936 937#    C   C   GQ=123;MQ=60    0.344   1   2   44  44
chrI    937 938#    G   G   GQ=10;MQ=60 0.197   8   33  44  44
chrI    955 956#    C   C   GQ=118;MQ=60    0.699   2   1   42  42

I have some downstream processing that relies on the bed files being well-formed, and would appreciate a solution.

Thanks,

Paul

[paul-sud]

I should add, this is using the most current gemBS revision.

[heathsc]

Hi Paul,

Thanks for the report - clearly a bug. Is this from the master or devel branch?

Cheers, Simon

On Mon, Jan 27, 2020 at 5:33 PM Paul Sud notifications@github.com wrote:

Reopened #69 https://github.com/heathsc/gemBS/issues/69.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/heathsc/gemBS/issues/69?email_source=notifications&email_token=AAY4652I3CIEI4ID7KVMCN3Q74EGHA5CNFSM4KMEGMNKYY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOWHCYTVA#event-2982513108, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAY4655HKDN27PQ2FK2BGYDQ74EGHANCNFSM4KMEGMNA .

[paul-sud]

Thanks for the quick response Simon. This is on master

[heathsc]

OK, in the file gemBS/tools/gemBS_plugins/output.c, line 133 is currently:

fprintf(fp,"%s\t%" PRId64 "\t%" PRId64 " \t%c", args->hdr->id [BCF_DT_CTG][rec->rid].key, rec->pos + pos, rec->pos + pos + 1, cx_len >= 3

• pos ? cx[2 + pos] : '.');

  and should be changed to:

fprintf(fp,"%s\t%" PRId64 "\t%" PRId64 "\t%c", args->hdr->id [BCF_DT_CTG][rec->rid].key, rec->pos + pos, rec->pos + pos + 1, cx_len >= 3

• pos ? cx[2 + pos] : '.');

If this solves your issue then I will push the fix to master.

Thanks again for the report. Simon

On Mon, Jan 27, 2020 at 5:42 PM Paul Sud notifications@github.com wrote:

Thanks for the quick response Simon. This is on master

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/heathsc/gemBS/issues/69?email_source=notifications&email_token=AAY46557C34P3VJVOOPOYQ3Q74FGNA5CNFSM4KMEGMNKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKAFY4A#issuecomment-578837616, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAY46522G23JAPQW5OTOQW3Q74FGNANCNFSM4KMEGMNA .

[heathsc]

I have just pushed the fix to master (and devel as the same fault was there as well)

On Mon, Jan 27, 2020 at 5:55 PM Simon Heath simon.heath@gmail.com wrote:

OK, in the file gemBS/tools/gemBS_plugins/output.c, line 133 is currently:

fprintf(fp,"%s\t%" PRId64 "\t%" PRId64 " \t%c", args->hdr->id [BCF_DT_CTG][rec->rid].key, rec->pos + pos, rec->pos + pos + 1, cx_len >= 3 + pos ? cx[2 + pos] : '.');

and should be changed to:

fprintf(fp,"%s\t%" PRId64 "\t%" PRId64 "\t%c", args->hdr->id [BCF_DT_CTG][rec->rid].key, rec->pos + pos, rec->pos + pos + 1, cx_len >= 3 + pos ? cx[2 + pos] : '.');

If this solves your issue then I will push the fix to master.

Thanks again for the report. Simon

On Mon, Jan 27, 2020 at 5:42 PM Paul Sud notifications@github.com wrote:

Thanks for the quick response Simon. This is on master

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/heathsc/gemBS/issues/69?email_source=notifications&email_token=AAY46557C34P3VJVOOPOYQ3Q74FGNA5CNFSM4KMEGMNKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKAFY4A#issuecomment-578837616, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAY46522G23JAPQW5OTOQW3Q74FGNANCNFSM4KMEGMNA .

[paul-sud]

Thanks Simon, the fix on master worked.