Matilda

[liuchunlei0430]

Thank you for submitting a method package to the HCA DCP Analysis Tools Registry!

To expedite your package's addition to https://data.humancellatlas.org/analyze/methods, please provide the following package metadata. You can easily edit this information later by clicking "Improve this page" at the bottom of your package's detail page (example).

Required:

• Tool title: Matilda

• Contact name: PYangLab

• Contact email: cliu@cmri.org.au

• Who to attribute: PYangLab

• Method-ready Docker image URL: https://github.com/PYangLab/Matilda/tree/main/img/main.jpg

• Code repository URL: https://github.com/PYangLab/Matilda

• Affirm code is free and open source with license available in code repository: [x] (fill in as [x] to affirm)

• Language: English

• Upstream registry URL: https://github.com/PYangLab/Matilda/tree/main/img/main.jpg

• Test data location: https://github.com/PYangLab/Matilda/tree/main/data/TEAseq

• Basic command line usage:

  # training the matilda model
  python main_matilda_train.py --rna [trainRNA] --adt [trainADT] --atac [trainATAC] --cty [traincty] #[training dataset]
  # python main_matilda_rna_train.py --rna [trainRNA] --cty [traincty]
  # Example run
  python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --adt ../data/TEAseq/train_adt.h5 --atac ../data/TEAseq/train_atac.h5 --cty ../data/TEAseq/train_cty.csv #Training TEAseq
  #python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --adt ../data/TEAseq/train_adt.h5  --cty ../data/TEAseq/train_cty.csv #Training CITEseq
  #python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --atac ../data/TEAseq/train_atac.h5 --cty ../data/TEAseq/train_cty.csv #Training RNA+ATAC
  # python main_matilda_rna_train.py --rna ../data/TEAseq/train_rna.h5 --cty ../data/TEAseq/train_cty.csv # for scRNA-seq

• Command to validate installation:

  # training the matilda model
  python main_matilda_train.py --rna [trainRNA] --adt [trainADT] --atac [trainATAC] --cty [traincty] #[training dataset]
  # python main_matilda_rna_train.py --rna [trainRNA] --cty [traincty]
  # Example run
  python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --adt ../data/TEAseq/train_adt.h5 --atac ../data/TEAseq/train_atac.h5 --cty ../data/TEAseq/train_cty.csv #Training TEAseq
  #python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --adt ../data/TEAseq/train_adt.h5  --cty ../data/TEAseq/train_cty.csv #Training CITEseq
  #python main_matilda_train.py --rna ../data/TEAseq/train_rna.h5 --atac ../data/TEAseq/train_atac.h5 --cty ../data/TEAseq/train_cty.csv #Training RNA+ATAC
  # python main_matilda_rna_train.py --rna ../data/TEAseq/train_rna.h5 --cty ../data/TEAseq/train_cty.csv # for scRNA-seq

• Short description: Matilda is a multi-task framework for learning from single-cell multimodal omics data. Matilda leverages the information from the multi-modality of such data and trains a neural network model to simultaneously learn multiple tasks including data simulation, dimension reduction, visualization, classification, and feature selection.

Optional:

• Long description: (will appear on detail page)
• Build badge URL: (e.g. https://travis-ci.org/kstreet13/slingshot.svg?branch=master)
• Coverage badge URL: (e.g. https://img.shields.io/codecov/c/github/kstreet13/slingshot/master.svg)
• Logo: (attach to this GitHub issue)
• Screenshot: (attach to this GitHub issue)
• Implementation of this tool in an active portal: URL
• Benchmark dataset: URL

[NoopDog]

@liuchunlei0430 this is live here https://data.humancellatlas.org/analyze/methods/matilda Thanks for submitting the package! Cheers, D