new feature request

[fdelsuc]

Hi,

Thank you for developing such a useful tool.

I have a suggestion for a new feature that I think would be easy to implement and useful in practice.
Would it be possible to specify a list of reference species that the initial random sampling of fragments is restricted to?
This would be particularly useful when analyzing datasets in which genome assemblies vary greatly in terms of contiguity and completeness. For example, it would be useful to restrict the initial random sampling of fragments to chromosome-scale reference genomes in a dataset that also includes much more fragmented genome assemblies.

Sincerely,

Frédéric

[ang037]

Hi Frédéric,

Glad to hear you found ROADIES useful, and thank you for suggesting this feature! We are currently working on an extended version of ROADIES that includes several new features, and this is one of the features we plan to release later this year. In the meantime, I am sharing instructions so you can use it right away if you would like.

This feature allows you to select specific input genomic files for sampling. Also, it gives you additional control over the proportion of gene samples from the selected genomes, rather than relying on purely random selection. This can be useful if, for example:

• You want some genomes to contribute more genes than others.
• Certain species groups are underrepresented, which could reduce signal if sampled purely at random.

By providing groups for each selected genome assembly, ROADIES will ensure that genes are sampled with equal probability across groups and, within each group, are evenly distributed among species.

Instructions

Step 1:

• Replace workflow/rules/sampling.smk with the attached file (make sure to change .txt to .smk)
  sampling.txt

• Replace workflow/Snakefile with the attached file (make sure to remove .txt - I added .txt since GitHub does not support attachements in other formats)
  Snakefile.txt

• Add the attached precompute_sampling.txt to the workflow/scripts folder.
  precompute_sampling.py

Step 2:

• Create a group_file.csv listing the genomes you want to select, along with their groups. If you’re unsure of groups, leave that field blank and end the line with a comma.
  An example group_file.csv is attached for reference.
  group_file.csv

Step 3:

• Add the following line to your config/config.yaml file:

GROUP_CSV: <path to your group_file.csv>

Replace <path to your group_file.csv> with the actual path to the file you created in Step 2.

Step 4:

• Run ROADIES as usual. It will now generate an extra file named sampling_output.txt (sample file attached). This file shows:
• Which genomic assemblies exist in the input directory
• Which genomic assemblies were selected
• The starting and ending gene IDs for each genome assembly

sampling_output.txt

Example: od = 1 and od_e = 250 means genes 1–250 were selected from that genome assembly.

IMPORTANT: Providing group_file.csv is optional. If you don’t provide it, ROADIES will perform random sampling as before.

We plan to include this functionality in the upcoming extended version of ROADIES. In the meantime, if anything is unclear or doesn’t work as expected, please let me know!

Regards,
Anshu

[fdelsuc]

Dear Anshu,

Thank you very much for your lightning-fast reply!
It's great to know that you are planning to release an extended version of ROADIES including several new features later this year. I am looking forward to it.
Thank you also for providing the files to implement my request so quickly. This is very helpful.
I followed your instructions, and the pipeline is now running with fragment sampling restricted to the species I declared in the 'group_file.csv' file, as reflected in the 'sampling_output.txt' file.
I will let you know if I encounter any problems with the new configuration, but so far, so good!

Best wishes,
Frederic