Transformer-based neural consensus caller for Pacific Biosciences HiFi sequencing
DeepConsensus applies deep learning to correct errors in Pacific Biosciences (PacBio) circular consensus sequencing (CCS), significantly improving yield of Q30+ and Q40+ high-accuracy reads.
Deep learning consensus calling across subreads to correct systematic sequencing errorsIncreases the yield of Q30 (>99.9% accurate) and Q40 (>99.99% accurate) HiFi reads by up to 40%Integrated into standard PacBio SMRT Link secondary analysis pipelinesOptimized with GPU acceleration for high-throughput sequencing facilities
DeepConsensus applies deep learning to correct errors in Pacific Biosciences (PacBio) circular consensus sequencing (CCS), significantly improving yield of Q30+ and Q40+ high-accuracy reads.
Information checked against an official source; not a hands-on test. Source · Last reviewed: 20/09/2026, 11:09:40
Key Features
Deep learning consensus calling across subreads to correct systematic sequencing errors
Increases the yield of Q30 (>99.9% accurate) and Q40 (>99.99% accurate) HiFi reads by up to 40%
Integrated into standard PacBio SMRT Link secondary analysis pipelines
Optimized with GPU acceleration for high-throughput sequencing facilities
Academic Context & Research Evidence
Biological & Workflow Fit
Biological Application
De novo genome assembly, structural variant discovery, and pangenomics
Research Workflow
Raw PacBio subreads -> DeepConsensus inference -> high-accuracy HiFi BAM/FASTQ
Use this citation format when referencing DeepConsensus in scientific publications and benchmark papers.
@software{deepconsensus_2026,
title = {{DeepConsensus}},
author = {{Google Health & PacBio}},
year = {2026},
url = {https://github.com/google/deepconsensus},
note = {Indexed on aibioatlas - AI for Biology and Drug Discovery}
}
Peer-Reviewed Literature & Preprints
Live scientific citations streamed from Europe PMC and PubMed for DeepConsensus.
⏳ Fetching real-time literature from Europe PMC & PubMed...
Technical / Product Information
Missing values mean the catalog has no recorded information. They do not mean a feature is absent.
Entry typeSoftware Tool
Access modeOpen Source
AI roleSignal & Sequence Error Correction
Input dataNot recorded
Output dataNot recorded
Licence conditionsBSD-3-Clause
Commercial eligibilityOpen source release
Compute requirementsMulti-core CPU or NVIDIA GPU (CUDA)
ValidationNot recorded
TypeNeural sequencing consensus caller
Intended useNot recorded
CompatibilityNot recorded
ManufacturerGoogle LLC & PacBio
Biological applicationDe novo genome assembly, structural variant discovery, and pangenomics