How are you! I met a problem when I use genome-guided Trinity on Galaxy server to assemble transscriptome. Below is how I operated:
Enter “Trinity” in the search box located at the upper left corner, then click on “Trinity” to enter the settings interface. The specific settings are shown below.
Tool Parameters Are you pooling sequence datasets? → seleted Yes; Paired or Single-end data? → selected Paired-end Left/Forward strand reads * → selected my trimmomatic-treated clean fastq.gz data Right/Reverse strand reads* → selected my trimmomatic-treated clean fastq.gz data Strand specific data → default No Jaccard Clip options → default No Run in silico normalization of reads → default Yes
Additional Options Minimum Contig Length - optional → default 200 Use the genome guided mode? → selected Yes Coordinate-sorted BAM file * → selected my STAR(software)-mapped, samtools sorted “bam” files Maximum allowed intron length (also maximum fragment span on genome) → type in 50000 Minimum read coverage for identifying an expressed region of the genome → default 1
“Minimum number of reads per partition” → default 10 Error-corrected or circular consensus (CCS) pac bio reads - optional → Nothing selected since my data was illumina reads. Minimum count for K-mers to be assembled - optional → default “1”
Finally, click on Run Tool.
Then an error appeared. Open the window of Tool Standard Output, it displays “Error - must specify either gene guided bam or fastq files, but not both. Use bam for gene guided and fastq for gene free de novo assembly. at/usr/local/bin/Trinity line 1148”. Since I want to used the genome-guided mode of Trinity, fastq files should not be provided as the message suggested. However, I found that “Left/Forward strand reads " and "Right/Reverse strand reads” are required and you will be reminded to upload these files, otherwise be left at this webpage. I have noticed the following help and answers, but Trinity at the Galaxy server can not run normally with both the FASTQ and BAM.
How could I operate to only upload Bam files without uploading fastq.gz files, or to run Trinity normally with both the FASTQ and BAM. Thanks in advance!
Your error message is odd, and is from the underlying tool so may just be a default warning that doesn’t apply to the Galaxy wrapper (sometimes those cannot be suppressed).
Did you do the read mapping in Galaxy? With the same fastq files? This used to work so I’m not sure why it isn’t now.
Hi @luweidong
it seems (the latest version of) Trinity has an issue with data provided in BAM files. Many of my test jobs failed with similar message (either fastq or BAM), but Trinity requires FASTQ file(s). I have not tested older versions.
Maybe try StringTie/StringTie Merge: create gene models from individual samples using StringTie and merge individual models using STringTie Merge.
Kind regards,
Igor
Thank you for your quick responses and suggestions.
As @jennaj suggested, now I am using genome-guided Trinity assembly (Galaxy Version 2.9.1+galaxy2), not the latest version. So far, no error message appears although the program has not yet completed the analysis. So, I suspect that using the latest version of Trinity might be the main cause for this issue.
Yes! @igor, I have previously used StringTie/StringTie Merge on Galaxy server to assemble transcripts for my 12 samples, and these tools works normally. This time, I want to compare the results made by those tools.
Hi @luweidong Ok, glad to know the prior version worked! The best advice is to use that one, at least for now.
And thanks @igor for doing the testing! This seems familiar … but maybe is new. I couldn’t find an open ticket about the issue so maybe missed it or the update is still pending .
.