Galt
Let me get back to you on 'libs' shortly. Could I ask first about a bedToBigBed
error message?
The input bed file is 1,013,372 records, and has no track line header. Records
are of this format:
chr1 137005859 137005861 d_2_c
I sorted the bed file following bedToBigBed's warning -
sort -k1,1 -k2,2n $BED > $SORTEDBED
Then,
{xhost06}/projects/morgen/wtss_assembly/resources/tass_db_2>
./tass_db_2_bedToBigBed.sh
pass1 - making usageList (32 chroms): 834 millis
pass2 - checking and writing primary data (1013372 records, 4 fields): 2799
millis
index write: 1 millis
bedToBigBed: bedToBigBed.c:394: writeReducedOnceReturnReducedTwice: Assertion
`boundsPt == boundsEnd' failed.
./tass_db_2_bedToBigBed.sh: line 13: 14201 Aborted bedToBigBed
$SORTEDBED $SIZES $DIR/my.bb
At the point where the error was reported, the bigBed output file was 5.1 MB.
Is the error likely due to a problem in my bed file?
Thanks for your help -
Gordon
On 2010-02-01, at 1:04 PM, Galt Barber wrote:
>
> Very close there.
> Please try "make libs" instead of "make lib".
>
> Then you should be able to get raSqlQuery etc to compile.
>
> -Galt
>
> Gordon Robertson wrote:
>> Thanks for your quick response, Galt.
>> I changed 'X*Y' to '(X*Y)', exported what worked in the past -
>> export MACHTYPE=x86_64
>> export MYSQLINC=-I/usr/include/mysql
>> export MYSQLLIBS='-L/usr/lib64/mysql -lmysqlclient -lz -lcrypt -lnsl -lm
>> -L/usr/lib64 -lssl -lcrypto'
>> ...and ran 'make' from the src/utils folder. Many (!) tools compiled. The
>> following error was reported, probably because I've not done all earlier
>> steps correctly -
>> make[1]: *** No rule to make target `../../lib/x86_64/jkhgap.a', needed by
>> `raSqlQuery'. Stop.
>> make[1]: Leaving directory
>> `/home/grobertson/linux_x86_64/KentSourceTree/kent/src/utils/raSqlQuery'
>> make: *** [all] Error 2
>> Could you clarify this error? I'd compiled lib -
>> {xhost06}~/linux_x86_64/KentSourceTree/kent/src> make lib
>> make: Nothing to be done for `lib'.
>> ...but jkjgap.a is not in this folder -
>> {xhost06}~/linux_x86_64/KentSourceTree/kent/src/lib/x86_64> ll
>> total 6284
>> -rw-rw-r-- 1 grobertson users 893 Jan 30 18:24 jkOwnLib.a
>> -rw-rw-r-- 1 grobertson users 5370 Jan 30 18:24 jkweb.a
>> -rwxrwxr-x 1 grobertson users 0 Dec 1 2006 placeHolder.c
>> ---
>> I now have bedToBigBed, which was the original goal -
>> {xhost06}~/linux_x86_64/KentSourceTree/kent/src/utils/bedToBigBed>
>> bedToBigBed ./test.bed ./chrom.sizes ./out.bb pass1 - making usageList (3
>> chroms): 21 millis
>> pass2 - checking and writing primary data (12 records, 4 fields): 1 millis
>> index write: 0 millis
>> pass3 - writeReducedOnceReturnReducedTwice: 1 millis
>> further reductions: 0 millis
>> Thanks again.
>> G
>> On 2010-02-01, at 12:00 PM, [email protected] wrote:
>>> Message: 4
>>> Date: Mon, 01 Feb 2010 11:20:38 -0800
>>> From: Galt Barber <[email protected]>
>>> Subject: Re: [Genome] can't compile bedToBigBed on RHELinux 4.x?
>>> To: Gordon Robertson <[email protected]>
>>> Cc: [email protected]
>>> Message-ID: <[email protected]>
>>> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>>>
>>> Hi, Gordon!
>>>
>>> Looks like you are using gcc 3.4
>>> and we are using gcc 4.1.2 on 64-bit linux.
>>>
>>> Either the gcc3.4 is more sensitive
>>> and gcc4 has a bug in not showing the
>>> warning; or, perhaps the gcc developers
>>> decided that this shouldn't be a warning after all.
>>>
>>> In any case, simply adding parentheses around the
>>> argument made both compilers happy.
>>>
>>> bits64 is definted as unsigned long long.
>>>
>>> In effect we had this:
>>>
>>> printf("%llu", (bits64)X*Y);
>>>
>>> and both versions of gcc are happy when we changed it to
>>>
>>> printf("%llu", (bits64)(X*Y));
>>>
>>> This change has been commmitted.
>>> You can either update this file from CVS
>>> or else you can just add the parentheses yourself.
>>>
>>> -Galt
>>>
>>> Gordon Robertson wrote:
>>>> Is there a problem with the current Kent source tree bedToBigBed?
>>>> I downloaded jksrc.zip tonight to 64-bit Red Hat Enterprise Linux 4.x. I'd
>>>> been using wigToBigWig for some time on that machine, but wanted to ensure
>>>> that I was using the latest version. I recompiled it after cd'ing to the
>>>> src/utils/wigToBigWig folder -
>>>> {xhost06}~/linux_x86_64/KentSourceTree/kent/src/utils/wigToBigWig> make
>>>> gcc -O -g -Wall -Werror -Wformat -Wimplicit -Wreturn-type -Wuninitialized
>>>> -o /home/grobertson/bin/x86_64/wigToBigWig wigToBigWig.o
>>>> ../../lib/x86_64/jkweb.a -lm -lz
>>>> {xhost06}~/linux_x86_64/KentSourceTree/kent/src/utils/wigToBigWig>
>>>> wigToBigWig
>>>> wigToBigWig v 3 - Convert ascii format wig file (in fixedStep, variableStep
>>>> or bedGraph format) to binary big wig format.
>>>> ...
>>>>
>>>> I want to start to use bedToBigBed. When I ran make from the src/utils
>>>> folder, a number of folders were processed (aNotB, addCols, ave, aveCols,
>>>> bedClip, bedGraphToBigWig) but there was an error message for bedToBigBed
>>>> -
>>>> make[1]: Entering directory
>>>> `/home/grobertson/linux_x86_64/KentSourceTree/kent/src/utils/bedToBigBed'
>>>> gcc -O -g -Wall -Werror -Wformat -Wimplicit -Wreturn-type -Wuninitialized
>>>> -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_GNU_SOURCE -DMACHTYPE_x86_64
>>>> -Wall -Werror -Wformat -Wimplicit -Wreturn-type -Wuninitialized -I../inc
>>>> -I../../inc -I../../../inc -I../../../../inc -I../../../../../inc -o
>>>> bedToBigBed.o -c bedToBigBed.c
>>>> bedToBigBed.c: In function `bbFileCreate':
>>>> bedToBigBed.c:537: warning: long long unsigned int format, long unsigned
>>>> int arg (arg 4)
>>>> make[1]: *** [bedToBigBed.o] Error 1
>>>> make[1]: Leaving directory
>>>> `/home/grobertson/linux_x86_64/KentSourceTree/kent/src/utils/bedToBigBed'
>>>> make: *** [all] Error 2
>>>>
>>>> If I'm missing something simple I apologize for wasting the list's time.
>>>> Thanks for your help with this.
>>>>
>>>> Gordon
>> --
>> Gordon Robertson
>> BC Cancer Agency Genome Sciences Centre
>> Vancouver BC Canada
>> _______________________________________________
>> Genome maillist - [email protected]
>> https://lists.soe.ucsc.edu/mailman/listinfo/genome
--
Gordon Robertson
BC Cancer Agency Genome Sciences Centre
Vancouver BC Canada
_______________________________________________
Genome maillist - [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome
