Skip to main content

Detecting novel pathogens from NGS reads in real-time during a sequencing run.

Project description

DeePaC-Live

A DeePaC plugin for real-time analysis of Illumina sequencing runs. Captures HiLive2 output and uses deep neural nets to detect novel pathogens directly from NGS reads.

We recommend having a look at:

Installation

# Optional, but recommended: for GPU users
conda install tensorflow-gpu
# Install deepac-live
conda install -c bioconda deepac-live
# Recommended: download and compile deepac-live custom models
deepac getmodels --fetch
# Optional: viral built-in models
conda install -c bioconda deepacvir

Alternatively, you can also use pip:

# Optional, but recommended: for GPU users
pip install tensorflow-gpu
# Install deepac-live
pip install deepac-live
# Recommended: download and compile deepac-live custom models
deepac getmodels --fetch

# Optional: viral built-in models (not necessary)
pip install deepacvir

DeePaC-Live models

DeePaC-Live ships new, updated models for bacterial pathogenic potential and viral infectious potential prediction. The Illumina models are trained on 25-250bp subreads to ensure high performance over the whole sequencing run. The Nanopore models are trained on 250bp subreads corresponding to just around 0.5s of sequencing. To fetch the models, use deepac getmodels --fetch. In the created directory, you will find the following models ready for inference:

  • illu-bac-res18.h5 : an Illumina bacterial model
  • illu-vir-res18.h5 : an Illumina viral model
  • nano-bac-res18.h5 : a Nanopore bacterial model
  • illu-vir-res18.h5 : a Nanopore viral model

Basic usage

# Run locally: deepac-live Illumina models
deepac-live local -C -m illu-bac-res18.h5 -s 25,50,75,100,133,158,183,208 -l 100 -i hilive-out -o temp -I temp -O output -B ACAG-TCGA,undetermined
# Run locally: custom model
deepac-live local -C -m custom_model.h5 -s 25,50,75,100,133,158,183,208 -l 100 -i hilive-out -o temp -I temp -O output -B ACAG-TCGA,undetermined

# Run locally: build-in model for bacteria (not recommended)
deepac-live local -c deepac -m rapid -s 25,50,75,100,133,158,183,208 -l 100 -i hilive-out -o temp -I temp -O output -B ACAG-TCGA,undetermined
# Run locally: build-in model for viruses (not recommended)
deepac-live local -c deepacvir -m rapid -s 25,50,75,100,133,158,183,208 -l 100 -i hilive-out -o temp -I temp -O output -B ACAG-TCGA,undetermined

Advanced usage

Setting up a remote receiver

# Setup sender on the source machine
deepac-live sender -s 25,50,75,100,133,158,183,208 -l 100 -A -i hilive-out -o temp -r user@remote.host:~/rem-temp -k privatekey -B ACAG-TCGA,undetermined
# Setup receiver on the target machine
deepac-live receiver -c deepacvir -m rapid -s 25,50,75,100,133,158,183,208 -l 100 -I rem-temp -O output -B ACAG-TCGA,undetermined

Refilter: ensembles and alternative thresholds

# Setup an ensemble on the target machine
deepac-live refilter -s 25,50,75,100,133,158,183,208 -l 100 -i rem-temp -I output_1,output_2 -O final_output -B ACAG-TCGA,undetermined
# Use another threshold
deepac-live refilter -s 25,50,75,100,133,158,183,208 -l 100 -i rem-temp -I output_1 -O final_output -t 0.75 -B ACAG-TCGA,undetermined

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

deepaclive-0.3.2.tar.gz (12.0 kB view details)

Uploaded Source

Built Distribution

deepaclive-0.3.2-py3-none-any.whl (13.8 kB view details)

Uploaded Python 3

File details

Details for the file deepaclive-0.3.2.tar.gz.

File metadata

  • Download URL: deepaclive-0.3.2.tar.gz
  • Upload date:
  • Size: 12.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/49.6.0.post20201009 requests-toolbelt/0.9.1 tqdm/4.50.2 CPython/3.8.6

File hashes

Hashes for deepaclive-0.3.2.tar.gz
Algorithm Hash digest
SHA256 aa69f4967710787f12694ac146b6ed07a78e876a93626b2fd1904efd03e8dde5
MD5 aff243d803c32f38a9a166cf9805c3fd
BLAKE2b-256 54bf7583e9990a187dc834b393981884a1e7e4515535ebe60ba545ae43809413

See more details on using hashes here.

File details

Details for the file deepaclive-0.3.2-py3-none-any.whl.

File metadata

  • Download URL: deepaclive-0.3.2-py3-none-any.whl
  • Upload date:
  • Size: 13.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/49.6.0.post20201009 requests-toolbelt/0.9.1 tqdm/4.50.2 CPython/3.8.6

File hashes

Hashes for deepaclive-0.3.2-py3-none-any.whl
Algorithm Hash digest
SHA256 272f02cefdaf2b79d555fe3b41638c8c849f3ea06971a9db3d4dc484ccb14315
MD5 cd6a1030024355e0f1bca55b8eaf21b2
BLAKE2b-256 fd65b2512ec98ef6f5cc1173799b1d0236e19d710dada821870404b6e3cbb467

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page