Questions before starting...

[ascialla]

Hello there,
I have a few questions before I start using the tool on my data.

I installed NanoRepeat and ran the tool on the data as described in the documentation, and everything went well.

My questions are as follows:

1. When providing fastq files, will the tool perform the alignment before proceeding to identify the STR targets? Since my data is WGS, I think the tool will take days to perform the alignment and consume many more resources.

Not having access to an HPC, I wanted to try to understand how to proceed first. In my case, I would use a PromethION 24.

Best regards and thanks in advance.

P.S. it seems that alignment is mandatory and a good practice to optimize time and resources.

[fangli80]

Hello,

Yes, you are correct. Alignment is the first step. Aligning a whole genome sequencing (WGS) dataset to a human reference can take several days without access to a High-Performance Computing (HPC) cluster.

If you are strictly interested in specific genes or regions, you can use a "bait-and-capture" approach to save time. Here is the workflow:

1. Create a subset reference: Create a small .fasta file that contains only the regions or genes of interest (you can extract these from the full reference using samtools faidx).
2. Initial alignment: Align all your raw reads to this small reference file using minimap2. This acts as a filter and will be significantly faster than mapping to the whole genome.
3. Extract mapped reads: Use samtools to grab only the reads that successfully aligned to your target regions.
   • Command: samtools fastq -F 4 input.bam > mapped.fastq
   • (Note: The -F 4 flag excludes unmapped reads).
4. Re-alignment (Validation): Take the resulting mapped.fastq and align it to the full reference genome (e.g., hg38).
5. Analysis: Use NanoRepeat to identify STRs using the new BAM file generated in step 4. （input type is bam instead of fastq）

Important Note: The re-alignment in Step 4 is critical. In Step 2, the aligner is forced to fit reads to your small reference even if they don't belong there (e.g., they might belong to a similar-looking region elsewhere in the genome). Re-aligning to the full genome ensures that the reads actually originate from your target region, preventing false positives.

[ascialla]

Hi fangli80,
Thank you for your complete and comprehensive reply.

Regarding step 5, I would like to identify the target sequences, some of which are interspersed with other repeated sequences, for example the HTT gene. What is your advice in this case? Which script should I run? nanoReapeat joint?

Or should we take a different approach to these target sequences?

Regards.

[fangli80]

Regarding the HTT gene, I recommend using the nanoRepeat-joint.py script. This script is specifically designed for analyzing regions where two tandem repeats are adjacent to each other.

For simple tandem repeat regions that do not have other tandem repeats nearby, please use nanoRepeat.py instead.