google / deepconsensus

DeepConsensus uses gap-aware sequence transformers to correct errors in Pacific Biosciences (PacBio) Circular Consensus Sequencing (CCS) data.
BSD 3-Clause "New" or "Revised" License
222 stars 37 forks source link

Nofound model for Quick start #11

Closed xuzaoxu closed 2 years ago

xuzaoxu commented 2 years ago

I followed "Quick start for DeepConsensus", downloaded test data. However, when I run commend: gsutil cp gs://brain-genomics-public/research/deepconsensus/models/v0.1/* "${MODEL_DIR}"/ There is no no matches found.

I also wonder what does "checkpoint-50" mean for "python3 -m deepconsensus.scripts.run_deepconsensus". How to get checkpoint file if I want to run my own data?

xuzaoxu commented 2 years ago

I used someone else's computer download ”gsutil cp gs://brain-genomics-public/research/deepconsensus/models/v0.1/*“ successfully. My question is checkpoint-50 applicable to different species data?

AndrewCarroll commented 2 years ago

Hi @xuzaoxu

Our intention is for one DeepConsensus model to work across any species. We have evaluated DeepConsensus on a non-human species (Z. mays) and observed improved HiFi results with downstream assembly methods. The generalization of DeepConsensus across species is something that we will monitor, with the goal of a single, universal model for PacBio HiFi error correction.