Software ecosystem

Software and resources

Our software ecosystem supports the lab's research programme in cell identity, interpretable AI, single-cell and spatial omics, organoid systems, disease-state modelling, method benchmarking, and biological fidelity assessment.

The resources below are organised to match the five research directions on the Research page, with a cross-cutting Benchmark tab for standard-setting work. General-purpose tools and legacy projects are retained in one additional tab.

Research narrative PYangLab GitHub

Organisation

Aligned to the five research directions.
Benchmarking, standard-setting and fidelity-assessment resources grouped separately.
Broad-use and legacy tools retained in one additional tab.
Links to code, documentation, and papers.

Research direction 1

Cell identity and cell-fate decisions

Tools for defining cell identity, improving annotation, integrating single-cell references, modelling multimodal identity signals and prioritising candidate regulators or compounds for cell-state conversion.

Cell identity genes

Cepo

Identifies cell-identity genes using differential stability in single-cell RNA-seq data.

Docs Bioconductor Paper

Annotation correction

scReClassify

Post hoc cell-type reclassification method for correcting potentially mislabelled single-cell annotations.

Repo Bioconductor Paper

Multiscale annotation

scClassify

Performs multiscale cell-type classification using single or multiple references and supports sample-size estimation.

Docs Bioconductor Paper

Single-cell integration

scMerge

Integrates and normalises single-cell RNA-seq data using stably expressed genes and pseudo-replicates.

Docs Bioconductor Paper

Multimodal identity modelling

Matilda

Multi-task learning framework for multimodal single-cell omics, including simulation, dimension reduction, classification and feature selection.

Repo Docs PyPI Paper

Cell-state conversion

Refate

Prioritises candidate genes, networks and compounds for directed cellular conversion from starting to target cell states.

Repo Tutorial Shiny Preprint

Research direction 2

Interpretable AI for single-cell systems biology

Tools and frameworks for learning from high-dimensional single-cell and multimodal data while exposing the features, programs and cell states that support each prediction.

Interpretable deep generative learning

Hydra

Deep generative ensemble learning framework for unimodal and multimodal single-cell omics, with interpretable feature attribution and cell-type annotation.

Repo Tutorial PyPI Paper

Condition-aware generative learning

M3

Deep generative framework for condition-aware multimodal single-cell integration, patient-level inference and multi-resolution attribution.

Repo Docs

Multi-task learning

Matilda

Jointly learns several multimodal single-cell tasks, including simulation, classification, feature selection and representation learning.

Repo Docs PyPI Paper

Feature selection benchmark

scDeepFeatures

Evaluation resource for deep learning-based feature selection in single-cell RNA-seq analysis.

Repo Paper

Positive-unlabelled learning

AdaSampling

Adaptive sampling for positive-unlabelled and label-noise learning in bioinformatics applications.

Repo CRAN Paper

Research direction 3

Single-cell, spatial and multimodal omics tools

Tools for data integration, annotation, multimodal analysis, phosphoproteomics and related machine-learning foundations.

CITE-seq analysis

CiteFuse

Supports preprocessing, integration, clustering, differential analysis and visualisation for CITE-seq data.

Docs Bioconductor Paper

Phosphoproteomics

PhosR

Processes phosphoproteomic data and supports kinase, pathway, signalome and phosphorylation-site analyses.

Docs Bioconductor Paper Protocol

Knowledge-guided clustering

CLUEY

Knowledge-guided framework for cell-type detection and clustering of unimodal and multimodal single-cell omics data.

Repo Paper PMC

Reference stable features

Stable genes and phosphosites

Reference stable genes and phosphorylation sites for single-cell and phosphoproteomic normalisation and integration.

Mouse SEGs Human SEGs Human SPSs

Phosphoproteomic clustering

ClueR

Knowledge-based clustering for phosphoproteomic time-series data using kinase-substrate relationships.

Repo CRAN Paper

Pathway and kinase analysis

directPA and KinasePA

Pathway and kinase perturbation analysis for transcriptomic, proteomic and phosphoproteomic datasets.

Repo CRAN directPA KinasePA

Kinase-substrate prediction

KSP-PUEL

Positive-unlabelled ensemble learning for kinase-substrate prediction from dynamic phosphoproteomics.

Repo Paper

Cross-cutting impact

Benchmarking, standards and fidelity assessment

Method benchmarks and biological fidelity resources that help the field move from plausibility-based assessment towards quantitative, task-specific and biologically grounded standards.

Multimodal integration standards

scMultiBench

Multitask benchmark for single-cell multimodal omics integration methods across vertical, diagonal, mosaic and cross-modality settings.

Repo Shiny Paper Highlight

Spatial variable gene benchmarking

SVGbench

Benchmark resource for evaluating spatially variable gene detection methods in spatial transcriptomics data.

Repo Paper

Simulation benchmarking

SimBench

Evaluation framework for comparing single-cell RNA-seq simulation methods by data-property estimation, biological-signal retention, scalability and applicability.

Repo Paper

Cell-type-number estimation

scCCESS

Consensus-clustering framework and benchmark resource for estimating the number of cell types in single-cell RNA-seq data.

Repo Benchmark paper Method paper

Feature-selection evaluation

scDeepFeatures

Evaluation resource for deep learning-based feature selection in single-cell RNA-seq analysis.

Repo Paper

Retinal organoid fidelity

Eikon

Atlas-linked resource for assessing retinal cell identity and the fidelity of human retinal organoids.

Shiny Paper

Brain culture fidelity

BrainSTEM

Single-cell fetal brain atlas resource for assessing transcriptomic fidelity of human midbrain cultures.

Resource Paper

Pluripotency reference

Stem Cell Atlas

Multi-omic reference atlas for benchmarking phased progression of pluripotency and stem-cell state transitions.

Resource Paper

Research direction 4

Stem cell and organoid systems

Resources for assessing cell-state fidelity, developmental progression and disease-relevant organoid systems.

Retinal organoids

Eikon

Interactive retinal cell identity and retinal organoid fidelity resource.

Shiny Paper

Brain atlas

BrainSTEM

Single-cell multiresolution fetal brain atlas for culture and organoid fidelity assessment.

Resource Paper

Pluripotency atlas

Stem Cell Atlas

Multi-omic resource for phased progression of pluripotency.

Resource Paper

Cerebral organoids

TransOmicsData

Data and code resource for trans-omic profiling of early human cerebral organoid formation.

Repo Paper

Research direction 5

Disease-state modelling and therapeutic prioritisation

Tools for identifying disease-associated cell states, interpreting altered molecular features and prioritising candidate targets or compounds.

Disease-associated cells

Hydra

Interpretable deep generative ensemble learning framework for identifying cell-type-specific features and preserving disease-relevant signatures.

Repo Tutorial PyPI Paper

Directional molecular change

SwitchClass

Label-switch classification framework for distinguishing attenuated and escalated molecular features across perturbation, recovery and disease trajectories.

Repo Preprint

Sample-level modelling

scFeatures

Generates sample-level multi-view representations from single-cell and spatial data for disease-outcome modelling.

Docs Bioconductor Paper

Candidate intervention prioritisation

Refate

Links regulatory-network modelling to candidate compound prioritisation for directed cell-state conversion and reprogramming hypotheses.

Repo Tutorial Shiny Preprint

Signalling and phosphorylation

PhosR

Supports phosphoproteomic processing and kinase/signalling analyses relevant to disease-state interpretation.

Docs Bioconductor Paper

General-purpose tools

General purpose machine learning tools

AdaSampling

Semi-supervised adaptive sampling for positive-unlabelled learning and learning from noisy class labels in bioinformatics applications.

Repo CRAN Paper

Sample Subset Optimization

Evolutionary sample-subset selection for imbalanced data and ensemble learning problems in bioinformatics applications.

Software TCYB paper PAKDD paper

Legacy projects

Archived Google Code projects

Historical software projects are kept here as part of the lab's research and methods record.

Re-Fraction

Machine-learning algorithm in Java for protein inference.

Archive

Self-Boosted Percolator

Boosted learning algorithm in Java for peptide filtering.

Archive

Genetic Ensemble SNPX

Parallel genetic algorithm in Java for SNP interaction detection.

Archive

Ensemble of Filters

Ensemble algorithm in Perl for SNP interaction filtering.

Archive

OCAP

Open source mass spectrometry analysis pipeline in R.

Archive

DyWave

Dynamic wavelet package in C/C++ for mass spectrum modelling.

Archive

Imbalanced Data Sampling

Particle swarm optimisation algorithm in Java for imbalanced data sampling.

Archive