Problem running JAFFA

[vincent...@gmail.com]

Hello!

First thanks for adding the tutorial for the generation of other annotations than the one provided.

I was trying the run test of JAFFA on a cluster running under linux

uname -a
Linux .... 2.6.32-504.30.3.el6.x86_64 #1 SMP Tue Jul 14 11:18:03 CDT 2015 x86_64 x86_64 x86_64 GNU/Linux

but got an error that I am not able to solve. Could you please help me?
NB: The R_libs folder contains the IRanges library compiled with R-3.1.3 and I am able to load it when running the loaded R module 3.1.0.

module load R/3.1.0-goolf-1.6.10
export R_LIBS="/home/ulg/genan/vhahaut/R_libs/"

cd jaffa-output/
$ ../JAFFA-version-1.06/tools/bin/bpipe run ../JAFFA-version-1.06/JAFFA_assembly.groovy ../demo_JAFFA/BT474-demo_1.fastq.gz ../demo_JAFFA/BT474-demo_2.fastq.gz
====================================================================================================
| Starting Pipeline at 2015-11-04 11:05 |
====================================================================================================

========================================= Stage run_check ==========================================
Running JAFFA version 1.06
Checking for required data files...
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19_genCode19.fa
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19_genCode19.tab
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/known_fusions.txt
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19.fa
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/Masked_hg19.1.bt2
/mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19_genCode19.1.bt2
All looking good

========================== Stage make_dir_using_fastq_names [BT474-demo] ===========================

================================= Stage prepare_reads [BT474-demo] =================================
TrimmomaticPE: Started with arguments: -threads 1 -phred33 ../demo_JAFFA/BT474-demo_1.fastq.gz ../demo_JAFFA/BT474-demo_2.fastq.gz BT474-demo/tempp1.fq /dev/null BT474-demo/tempp2.fq /dev/null LEADING:0 TRAILING:0 MINLEN:35
Input Read Pairs: 83014 Both Surviving: 83014 (100,00%) Forward Only Surviving: 0 (0,00%) Reverse Only Surviving: 0 (0,00%) Dropped: 0 (0,00%)
TrimmomaticPE: Completed successfully
83014 reads; of these:
83014 (100.00%) were paired; of these:
1527 (1.84%) aligned concordantly 0 times
81487 (98.16%) aligned concordantly exactly 1 time
0 (0.00%) aligned concordantly >1 times
98.16% overall alignment rate
1527 reads; of these:
1527 (100.00%) were paired; of these:
1347 (88.21%) aligned concordantly 0 times
180 (11.79%) aligned concordantly exactly 1 time
0 (0.00%) aligned concordantly >1 times
11.79% overall alignment rate

================================= Stage run_assembly [BT474-demo] ==================================
/mnt/fhgfs/Radek/JAFFA-version-1.06/tools/bin/velveth directory 19,36,4 -fastq -separate ../../BT474-demo/BT474-demo_filtered_reads.fastq.1.gz ../../BT474-demo/BT474-demo_filtered_reads.fastq.2.gz >> log_velveth
Running Assembly for k=19
Running Assembly for k=23
Running Assembly for k=27
Running Assembly for k=31
Running Assembly for k=35
Oases K-mer=19 assembly succeeded
Oases K-mer=23 assembly succeeded
Oases K-mer=27 assembly succeeded
Oases K-mer=31 assembly succeeded
Oases K-mer=35 assembly succeeded
Oases mergeAssembly succeeded
real 1m3.158s
user 0m44.400s
sys 0m3.496s

======================== Stage align_transcripts_to_annotation [BT474-demo] ========================
Loaded 195726195 letters in 99863 sequences
Searched 370731 bases in 236 sequences
real 0m10.929s
user 0m9.645s
sys 0m0.807s

============================== Stage filter_transcripts [BT474-demo] ===============================
Pipeline failed! (2)

Command failed with exit status = 1 :

time /usr/bin/R --vanilla --args BT474-demo/BT474-demo.psl BT474-demo/BT474-demo.txt 1000 /mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19_genCode19.tab < /mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/process_transcriptome_blat_table.R &> BT474-demo/log_filter

Pipeline failed! (2)

One or more parallel stages aborted. The following messages were reported:

Command failed with exit status = 1 :

time /usr/bin/R --vanilla --args BT474-demo/BT474-demo.psl BT474-demo/BT474-demo.txt 1000 /mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/hg19_genCode19.tab < /mnt/fhgfs/Radek/jaffa-output/../JAFFA-version-1.06/process_transcriptome_blat_table.R &> BT474-demo/log_filter

======================================== Pipeline Finished =========================================

Thanks in advance,

Vincent

[Nadia Davidson]

Hi Vincent,

Do you mind attaching or emailing me the file BT474-demo/log_filter?

Cheers,

Nadia.

[vincent...@gmail.com]

No need, I did not look carefully enough to this log file and the error is coming from R.

module load R/3.1.0-goolf-1.6.10

was not doing its job. I'll try again.

Thanks for your answer.

[Jack]

Hi Nadia,

I encounter the same problem.

The errors in "log_filter" file is the following:

------------------------------------------------------------------------------------
Warning messages:
1: In max(ends) : no non-missing arguments to max; returning -Inf
2: In min(starts) : no non-missing arguments to min; returning Inf
Error in do.call("rbind", lapply(split_results, multi_gene)) :
error in evaluating the argument 'args' in selecting a method for function 'do.call': Error in sapply(strsplit(genes, "__"), function(y) { :
error in evaluating the argument 'X' in selecting a method for function 'sapply': Error in strsplit(genes, "__") : non-character argument
Calls: lapply -> lapply -> FUN -> sapply
Execution halted
------------------------------------------------------------------------------------

Many thanks !

Jack

[Nadia Davidson]

Hi Jack,

Which version of R are you using? Can you also post the output from JAFFA and do an "ls -l" in the directory where you installed JAFFA and also the sample directory (e.g. BT474-demo).

Cheers,

Nadia.

[chich...@hotmail.com]

Hi Nadia,

The R version is 3.1.1 (2014-07-10)
I failed to download the demo file (e.g. BTF-474). Therefore, I ran the software using other data: HBRR2_1.fq.gz, HBRR2_2.fq.gz.

HBRR2_JAFFA.sh:
.../bpipe run -p fastqInputFormat="%_*.fq.gz" .../JAFFA_direct.groovy .../HBRR2_*.fq.gz

HBRR2_JAFFA.sh.o1679721:
================================= Stage filter_transcripts [HBRR2] =================================
Pipeline failed! (2)

Command failed with exit status = 1 :

time .../R-3.1.1/bin/R --vanilla --args HBRR2/HBRR2.psl HBRR2/HBRR2.txt 1000 .../JAFFA-version-1.07/hg19_genCode19.tab < .../JAFFA-version-1.07/process_transcriptome_blat_table.R &> HBRR2/log_filter

Pipeline failed! (2)

One or more parallel stages aborted. The following messages were reported:

Command failed with exit status = 1 :

time .../R-3.1.1/bin/R --vanilla --args HBRR2/HBRR2.psl HBRR2/HBRR2.txt 1000 .../JAFFA-version-1.07/hg19_genCode19.tab < .../JAFFA-version-1.07/process_transcriptome_blat_table.R &> HBRR2/log_filter
======================================== Pipeline Finished =========================================

The outputs in the directory 'HBRR2/' is the following:

6.5K Oct 10 14:46 log_filter
88 Oct 10 14:45 log_blat
159M Oct 10 14:45 HBRR2.psl
20M Oct 10 14:07 HBRR2.fasta
2.8M Oct 10 14:07 HBRR2_discordant_pairs.bam.bai
12M Oct 10 14:07 HBRR2_discordant_pairs.bam
215M Oct 10 14:04 HBRR2_filtered_reads.fastq.2.gz
213M Oct 10 14:04 HBRR2_filtered_reads.fastq.1.gz
9.9M Oct 10 14:04 HBRR2_leftover_reads.fastq.2.gz
9.8M Oct 10 14:04 HBRR2_leftover_reads.fastq.1.gz
0 Oct 10 13:14 HBRR2.ignore

The installation directory is the following:

22K Oct 9 20:50 JAFFA_stages.groovy
996 Oct 8 15:57 tools.groovy
306 Sep 21 14:50 tools
93 Sep 19 22:29 hg19.fa
2.3K Jun 24 13:13 assemble.sh
2.4K Jun 24 13:13 compile_results.R
1.1K Jun 24 13:13 get_spanning_reads_for_direct_1.R
999 Jun 24 13:13 get_spanning_reads_for_direct_2.R
1.6K Jun 24 13:13 get_spanning_reads.R
4.5K Jun 24 13:13 install_linux64.sh
1.3K Jun 24 13:13 JAFFA_assembly.groovy
1.5K Jun 24 13:13 JAFFA_direct.groovy
1.7K Jun 24 13:13 JAFFA_hybrid.groovy
14K Jun 24 13:13 make_final_table.R
4.4K Jun 24 13:13 process_transcriptome_blat_table.R
618 Jun 24 13:13 README
713 Jun 24 13:13 README.md
3.3G Mar 6 2015 JAFFA_REFERENCE_FILES_MM10_GENCODEVM4.tar.gz
25K Oct 23 2014 known_fusions.txt
668M Oct 22 2014 Masked_hg19.rev.2.bt2
898M Oct 22 2014 Masked_hg19.rev.1.bt2
668M Oct 22 2014 Masked_hg19.4.bt2
2.5M Oct 22 2014 Masked_hg19.3.bt2
668M Oct 22 2014 Masked_hg19.2.bt2
898M Oct 22 2014 Masked_hg19.1.bt2
80M Oct 20 2014 hg19_genCode19.rev.1.bt2
47M Oct 20 2014 hg19_genCode19.rev.2.bt2
80M Oct 20 2014 hg19_genCode19.1.bt2
47M Oct 20 2014 hg19_genCode19.2.bt2
24M Oct 20 2014 hg19_genCode19.tab
878K Oct 20 2014 hg19_genCode19.3.bt2
47M Oct 20 2014 hg19_genCode19.4.bt2
199M Oct 20 2014 hg19_genCode19.fa

Thanks,
Jack

[Nadia Davidson]

Hi Jack,

Thanks for this info. I can't see anything obvious that's out of place. The only time I've seen something similar to this was when someone accidentally called fusions for a mouse sample using human reference files.

Would you be able to share your file "HBRR2/HBRR2.psl" with me? either by google drive or dropbox? Or some similar example that I can use to reproduce the error.

Cheers,

Nadia.

[chich...@hotmail.com]

Hi Nadia,

I have provided the file "HBRR2" through google drive. You could download the file via your email "nadia.d...@mcri.edu.au".

Thanks,
Jack

[Nadia Davidson]

Thanks Jack,

I think I've spotted the problem. Your read IDs look like "FCC0TUMACXX:6:1101:1759:2102#/1/1" and R is treating the "#" like a comment and ignoring everything after it in the table of blat results. I'm going to fix this in future versions of JAFFA. However, I think there is another issue here too, there's an extra "/1" added to the end of the read IDs. This has probably been added by JAFFA somewhere along the line. I'm not 100% sure if this will matter, but there is the chance it could cause trouble with JAFFA being able to work out read pairs later on in the pipeline.

The simplest solution would be to modify the IDs in the input fastq files, removing the final character "#/1" or "#/2". You could do this using "sed" e.g. 

sed 's/#\/1//g' my_read1.fastq > my_new_read1.fastq

sed 's/#\/2//g' my_read2.fastq > my_new_read2.fastq

Let me know if you have any trouble and thanks for reporting the problem and sending some files through to reproduce it.

Cheers,

Nadia.

[maxmuelle...@gmail.com]

Hi,

I am new to JAFFA and want to try it for my master thesis about gene fusions.
I may have a similar problem while running the demo.
Jaffa stops in the filter_transcripts part:

"time /usr/bin/R --vanilla --args MCF7-demo/MCF7-demo.psl MCF7-demo/MCF7-demo.txt 1000 /media/mdhdd/richard/genfusion/test/JAFFA-version-1.09/hg38_genCode22.tab < /media/mdhdd/richard/genfusion/test/JAFFA-version-1.09/process_transcriptome_blat_table.R &> /media/mdhdd/richard/genfusion/test/result/MCF7-demo/log_filter"

If I look at the "../MCF7-demo/log_filter" I see the follwing error:
"> candidates=do.call("rbind",lapply(split_results,multi_gene))
Error in .local(query, subject, maxgap, minoverlap, type, select, ...) :
unused arguments (drop.self = TRUE, drop.redundant = TRUE)
Calls: do.call ... lapply -> lapply -> FUN -> findOverlaps -> findOverlaps"

The folder content looks like this:
97K Oct 6 14:42 MCF7-demo_leftover_reads.fastq.2.gz
93K Oct 6 14:42 MCF7-demo_leftover_reads.fastq.1.gz
5.5M Oct 6 14:42 MCF7-demo_filtered_reads.fastq.1.gz
5.6M Oct 6 14:42 MCF7-demo_filtered_reads.fastq.2.gz
312K Oct 6 14:43 MCF7-demo.fasta
4.0K Oct 6 14:43 oases
740K Oct 6 14:43 MCF7-demo.psl
84 Oct 6 14:43 log_blat
7.2K Oct 6 16:25 log_filter

Any help would be greatly appreciated.

Cheers,

[Nadia Davidson]

Hello,

I think this is to do with an incompatible version of JAFFA and R. Your options are either to upgrade your R/bioconductor to 3.5/3.4.0 or greater, or install JAFFA version 1.08 (https://github.com/Oshlack/JAFFA/releases) which should work for old versions of R.

Cheers,

Nadia.

[maxmuelle...@gmail.com]

Hi Nadja,

thank you for your reply.
I am already using "R version 3.4.0 (2017-04-21)".

Cheers,
Max

[Nadia Davidson]

Hi Max,

Does running JAFFA version 1.08 help?

Cheers,

Nadia.

[hog...@gmail.com]

I have the same problem as MAX.
I am running Jaffa-v1.09 and R-3.1.3

the below function gives follwoing error
lapply(split_results,multi_gene)
[1] 20

Error in .local(query, subject, maxgap, minoverlap, type, select, ...) :
unused arguments (drop.self = TRUE, drop.redundant = TRUE)

Any suggestions?

Thank you.

[Nadia Davidson]

Hi,

Did going back a version or two of JAFFA help you?

Cheers,

Nadia.