Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

pileUp and dmr showing significant results for m6A even with very low amount of mapped reads from Dorado's modBAM

Aperta
  • Qualcosa non va?
#618 2 commenti 0 reazioni 0 assegnatari Vedi su GitHub

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
4/5
Tempo stimato
3-5 giorni
Idoneità per principianti
35/100
Tipo di issue
Bug
Chiarezza
Da chiarire
Stato di attività
Tranquilla
Stack tecnologico
rust
Ambito
bioinformatics

Direzione di ricerca

Start with the modkit pileup and dmr entry points, using the reported modBAM inputs and the 0h, 24h, and 48h mapped-read counts. Reproduce the pass-threshold and missing-in-one-condition warnings, then compare the dmr bedMethyl results across read depths. Done means establishing whether the low-input 0h results are statistically valid and documenting or correcting the behavior.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

question

Dear modkit developers,

I've been facing this curious issue with m6A pileup) and differential methylation (dmr), and would like to hear your opinion on whether or not to trust on results:

I'm analysing m6A on a viral RNAs in a timeseries designh (0h, 24h, and 48h, of infection). On the 0h timepoint, expectedly, I've got very few reads aligned to the virus ref genome ( ~3k mapped reads only).
I know that modkit pileup recommends an N sampling of 10-50k reads for performing that "pass" threshold calculation (and I use -n 50000 on my runs), but it does still seem to run well even for these 0h samples like 0h_rep1 (~2.1k reads mapped) and 0h_rep3 (~4.8k reads mapped), finding a confidence 'pass' threshold over 80%.
Then, modkit dmr also seems to run well (on both 0h-24h and 0h-vs-48h), even though warning on a great amount of failed sites due to "missing-in-one-condition", which is the 0h one.

After filtering for pval < 0.05 (column 15 from the dmr bedMethyl), I get tens of differentially modified m6A sites, with an effect size (column) bias towards the 0h sample group. All sites are highly differentially modified in the 0h group.

Shall I trust these results? Or it's just a statistical bias from the little input size (%mapped reads) of the 0h sample.

JFYI: 24h and 48h have several hundreds thousands (200-700k) mapped reds against the same genome. And it's dmr bedMethyl file displays a good balance of effect sizes (some more modified in the 24h, and some more in the 48h).

Thanks,
Best,
Elton

Lingua principale
Rust
Stelle
274
Fork
33
Nessuna PR unita negli ultimi 30g

Preparare l'ambiente

Non abbiamo ancora controllato i file di configurazione di questo progetto. Parti dal suo README e consulta la nostra guida al primo contributo per i passaggi generali.

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di nanoporetech/modkit

Tutte le issue di nanoporetech/modkit

Issue simili

Altre issue su Rust

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.