Non-Single Isodecoders in isodecoder_counts_raw.txt from mim-tRNAseq

[Xinkai-Wu]

Hi, the maintainer of mim-tRNAseq,

I noticed that in the isodecoder_counts_raw.txt file, there are cases where non-single isodecoders appear. For example, the entry Homo_sapiens_tRNA-Ala-TGC-3/2/4.

From the isodecoder FASTA sequences, the corresponding entries are:

>Homo_sapiens_tRNA-Ala-TGC-3
GGGGATGTAGCTCAGTGGTAGAGCGCATGCTTTGCATGTATGAGGCCCCGGGTTCGATCCCCGGCATCTCCACCA  
>Homo_sapiens_tRNA-Ala-TGC-4
GGGGATGTAGCTCAGTGGTAGAGCGCATGCTTTGCACGTATGAGGCCCCGGGTTCAATCCCCGGCATCTCCACCA  
>Homo_sapiens_tRNA-Ala-TGC-2
GGGGATGTAGCTCAGTGGTAGAGCGCATGCTTTGCATGTATGAGGTCCCGGGTTCGATCCCCGGCATCTCCACCA  

Mim-tRNAseq assigns reads to different transcripts based on unique mismatch positions. However, there are still instances in the output file where multiple isodecoders are listed for the same entry, even though there are sequence differences between different isodecoder transcripts. However, in the mismatch.txt file, these references exist independently. Could you kindly clarify what might be causing this?

Best,
Kai

[Xinkai-Wu]

Additionally, if I want to combine the results of the CCAanalysis with those of DESeq2, can I simply match different isodecoders by their names? For example, does Gly-TCC-2-multi (from CCAanalysis) correspond to Homo_sapiens_tRNA-Gly-TCC-2/1 (from DESeq2)? In other words, how can I determine the correspondence between the two? Thanks.

[drewjbeh]

Hi there, Unique mismatches are one criteria for cluster deconvolution. However, if these unique mismatches overlap sites that also are likely modified, it becomes impossible to deconvolve these clusters because of misincorporation at these sites. Please see the section "Cluster deconvolution" and figure 4 in the protocols paper ( https://www.sciencedirect.com/science/article/pii/S2666166722004592?via%3Dihub#sec4 ). For this reason these Ala-TGC isodecoders are not deconvoluted. This deconvolution is consistent across all analyses, as long as you are looking within one run and not separate mimseq runs. You can see which transcripts are in which unsplit clusters and why they could not be deconvoluted by looking at the file 'unsplitClusterInfo.txt'

…

On Fri, 31 Jan 2025, 02:44 Xinkai, ***@***.***> wrote: Additionally, if I want to combine the results of the CCAanalysis with those of DESeq2, can I simply match different isodecoders by their names? For example, does Gly-TCC-2-multi (from CCAanalysis) correspond to Homo_sapiens_tRNA-Gly-TCC-2/1 (from DESeq2)? In other words, how can I determine the correspondence between the two? Thanks. — Reply to this email directly, view it on GitHub <#76 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACULZPOPMT3YGWV42BRO4U32NLIRNAVCNFSM6AAAAABUWR5LESVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRWGEYTAMBVHA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>