tiny-count does provide Feature.ID as Human genes?

[IperM]

Dear Authors,

I have recently been working with your pipeline. I attempted to install the entire pipeline, but due to some differences in the installation of Conda (micromamba is installed on the local machine), the installation was failing. Consequently, I decided to use tiny-count by itself instead.

For sake of using this tool, I used Bowtie (version 1.3.1) and Trimmomatic (version 0.39) locally for the preprocessing steps. Once this was completed, I placed the reference fasta and GTF files for the respective organisms (Macaca mulatta and Chlorocebus sabaeus).

When analyzing the values in R (version 4.4.2), I noticed that the Feature.ID correspond to Homo sapiens genes, suggesting that some gene conversion or synonym mapping is occurring between the genes from the original organisms and Homo sapiens.

Thank you for your time, and I appreciate the great work on your pipeline.

Ian

[AlexTate]

Hi @IperM, thanks for reaching out. The Feature ID is determined by the attributes in column 9 of your GTF files, and the order of precedence is ID > gene_id > Parent. tiny-count doesn't make any assumptions about the organism nor does it apply additional gene conversion or synonym mapping while determining the Feature ID. If you would like we can troubleshoot further, but I'll need to know the source of your GTF files and whether they have been modified, and I'll need a short sample of the Feature IDs that you've identified as Homo sapien genes.

[IperM]

After looking into the GTF file and going through the different steps with a more manual approach, I noticed that in the native GTF file downloaded from NCBI, some genes are already positioned the same way. Upon further investigation using different databases (GeneCards, _ENSEMBL_, and KEGG), I found that certain genes (e.g. MFAP2, CROCC, and SDHB) are listed with the same names in both organisms I'm working with (Homo sapiens and Chlorocebus sabaeus).

In principle, no further troubleshooting should be required. However, in case it's needed or for general interest, you can refer to the native GTF file for Vero available on NCBI.

Thanks again for everything — it was really helpful!