Can I use TRUST4 to deal with different length of reads?

[hz1010]

Hi!
TRUST4 is a great job. Now I have a my.fastq file in whose reads are in different length (eg. first read is 1000bp, second read is 1001bp). I wonder if I can use directly use TRUST4, or I need to trim those reads into same length?

Thanks

[mourisl]

Yes, TRUST4 can work with such fastq files. TRUST4 also works for the fastq files where read lengths are varied within one file.

[hz1010]

It's very nice of you. I still have a question, my fastq files are from Three-generation sequencing (by Nanopore). Do TRUST4 still work?

[mourisl]

I have tested with PacBio HiFi data and it works fine. For your Nanopore data, you need to correct the sequencing errors first.

[hz1010]

Thank you for your suggestion! But I have encountered an error when running TRUST4: failed: 256 at ./run-trust4 line 47. I don't know how to solve it and what it means.

[mourisl]

Can you show me the running log? What is the length of the longest read in your file?

[hz1010]

The length of the longest read in my file is 57798 bp. Running log:[Mon May 17 22:16:05 2021] TRUST4 begins.
[Mon May 17 22:16:05 2021] SYSTEM CALL: /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/fastq-extractor -t 1 -f /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/human_IMGT+C.fa -o test1_toassemble -u /xtdisk/jiangl_group/liyun/data/RNA-test/VDJ-ONT-10Xsp-sm_20210322/ONT/fastq_pass/PAG50586_pass_barcode05_3c2e0e42_0.fastq
[Mon May 17 22:16:05 2021] Start to extract candidate reads from read files.
[Mon May 17 22:16:12 2021] Finish extracting reads.
[Mon May 17 22:16:13 2021] SYSTEM CALL: /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/trust4 -f /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/human_IMGT+C.fa -o test1 -u test1_toassemble.fq
[Mon May 17 22:16:13 2021] SYSTEM CALL: /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/annotator -f /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/human_IMGT+C.fa -a test1_final.out -t 1 -o test1 -r test1_assembled_reads.fa > test1_annot.fa
Need to use -a to specify the assembly file.
system /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/annotator -f /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/human_IMGT+C.fa -a test1_final.out -t 1 -o test1 -r test1_assembled_reads.fa > test1_annot.fa failed: 256 at /xtdisk/jiangl_group/huangzh/software/TRUST4-1/TRUST4/run-trust4 line 47.

[mourisl]

TRUST4 can handle the read length up to 100K. It seems in your data, TRUST4 could not find any reads from the VDJ region, hence the file test1_final.out was empty, and the downstream annotation method failed. Just want to make sure, is your data corrected? With the error rate of raw Nanopore data, I don't think TRUST4 can identify the hit on V, J, C genes.

The read length also looks much longer than the gene sequence, is it from other genes or DNA-based sequencing?

[hz1010]

Thanks a lot for your help. It seems something wrong with my sequencing data,I will check my files again.

[hz1010]

Hello!
In my file, if VDJ regions repeated within one read, can TRUST4 still find them out? Or do I need to extract repeated regions out first before running TRUST4.
Thanks.

[mourisl]

If they are identical repeats, I think TRUST4 will pick the one that showed up first. If not, it may use V, J, C gene coordinates from different repeat structures, hence unable to identify CDR3. To be safe, I think you shall process the data first.