AstraZeneca/awesome-shapley-value

Stars
137
Rank 266,121 (Top 6 %)
Language
License
Apache License 2.0
Created over 2 years ago
Updated over 2 years ago

AstraZeneca/awesome-shapley-value

AstraZeneca

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Reading list for "The Shapley Value in Machine Learning" (JCAI 2022)

Awesome Shapley Value

The Survey Paper

This repository accompanies our survey paper The Shapley Value in Machine Learning.

If you find the survey or this repository useful in your research, please consider citing our paper:

@inproceedings{shapleysurvey,
    title     = {The Shapley Value in Machine Learning},
    author    = {Rozemberczki, Benedek and Watson, Lauren and Bayer, Péter and Yang, Hao-Tsung and Kiss, Olivér and Nilsson, Sebastian and Sarkar, Rik},
    booktitle = {Proceedings of the Thirty-First International Joint Conference on
               Artificial Intelligence, {IJCAI-22}},
    publisher = {International Joint Conferences on Artificial Intelligence Organization},
    pages     = {5572--5579},
    year      = {2022},
}

Contents

License

Apache 2.0

awesome-explainable-graph-reasoning

A collection of research papers and software related to explainability in graph machine learning.

chemicalx

A PyTorch and TorchDrug based deep learning library for drug pair scoring. (KDD 2022)

rexmex

A general purpose recommender metrics library for fair evaluation.

awesome-drug-discovery-knowledge-graphs

A collection of research papers, datasets and software related to knowledge graphs for drug discovery. Accompanies the paper "A review of biomedical datasets relating to drug discovery: a knowledge graph perspective" (Briefings in Bioinformatics, 2022)

SubTab

The official implementation of the paper, "SubTab: Subsetting Features of Tabular Data for Self-Supervised Representation Learning"

onto_merger

OntoMerger is an ontology alignment library for deduplicating knowledge graph nodes that represent the same domain.

awesome-drug-pair-scoring

Readings for "A Unified View of Relational Deep Learning for Drug Pair Scoring." (IJCAI 2022)

biology-for-ai

learning biology syllabus, geared for machine learning folks

KAZU

Fast, world class biomedical NER

jazzy

Fast calculation of hydrogen-bond strengths and free energy of hydration of small molecules.

judgyprophet

Forecasting for knowable future events using Bayesian informative priors (forecasting with judgmental-adjustment).

kallisto

Efficiently calculate 3D-features for quantitative structure-activity relationship approaches.

skywalkR

code for Gogleva et al manuscript

runnable

kgem-in-drug-discovery

Code to accompany the "Understanding the Performance of Knowledge Graph Embeddings in Drug Discovery" manuscript (Artificial Intelligence in the Life Sciences, 2022)

StarGazer

StarGazer is a tool designed for rapidly assessing drug repositioning opportunities. It combines multi-source, multi-omics data with a novel target prioritization scoring system in an interactive Python-based Streamlit dashboard. StarGazer displays target prioritization scores for genes associated with 1844 phenotypic traits.

Omicsfold

Multi-omics data normalisation, model fitting and visualisation.

VecNER

A library of tools for dictionary-based Named Entity Recognition (NER), based on word vector representations to expand dictionary terms.

peptide-tools

Programs to calculate phys-chem properties of synthetic peptides and proteins: isoelectric point and extinction coefficients.

data-science-python-course

Jupyter Notebook

napari-wsi

A plugin to read whole slide images within napari.

biomedical-kg-topological-imbalance

Code to accompany the "Implications of Topological Imbalance for Representation Learning on Biomedical Knowledge Graphs" (Briefings in Bioinformatics, 2022)

Jupyter Notebook

ibd-interpret

We trained high performing open source models on image scans of tissue biopsies to predict endoscopic categories in inflammatory bowel disease. These predictive models can help us better understand the disease pathology and represent a step towards automated clinical recruitment strategies.

roo

A package and environment manager for R

NESS

Official implementation of "NESS: Node Embeddings from Static Subgraphs"

detectIS

A pipeline to rapidly detect exogenous DNA integration sites using DNA or RNA paired-end sequencing data

skywalkR-graph-features

Example notebooks that illustrate how to generate knowledge-based features. Features can be used in a variety of ML models, including recommender systems.

Jupyter Notebook

multimodal-python-course

The purpose of the code is to facilitate a comprehensive understanding of multimodal data science applications within medical domain. The code serves to support the delivery of a cutting-edge workshop designed to introduce researchers to the rapidly evolving field of multimodal data science

Jupyter Notebook

UnlockingHeart

This repository accompanies our paper Unlocking the Heart Using Adaptive Locked Agnostic Networks and enables replication of the key results.

Tendril

This repository contains R package code for calculating tendril plots.

CTELC-Patient-Attrition-Model

Clinical Trial Enrollment Life Cycle (CTELC) modeling project aims to leverage "industry-wide" data to understand key drivers and build predictive models. Patient attrition, also referred to as dropout or patient withdrawal, occurs when patients enrolled in a clinical trial either withdraw or are lost to follow-up by the clinical site and trial sponsor.

MVDA_exploration_tools

Multivariate data analysis (MVDA) exploration tool is a Python library utilizing the scikit-learn library for partial least squares (PLS) and principal components analysis (PCA).

Jupyter Notebook

Siamese-Regression-Pairing

Siamese Neural Networks for Regression: Similarity-Based Pairing and Uncertainty Quantification

ConvCaps-DR

Tensorflow-Keras implementation of deep Convolutional Capsule Networks with Dynamic Routing algorithm

hsqc_structure_elucidation

Implementation of the SGNN graph neural network for 1H and 13C NMR prediction and a tool for distinguishing different molecules based on HSQC simulations

Jupyter Notebook

Multimodal_NSCLC

multi-omics data integration helps improving patient survival prediction. We provide a pipeline allowing for early integration of multiple omics plus clinical modalities in order to predict patient survival for NSCLC. The pipeline utilizes autoencoders, and helps identify main driving factor in survival prediction

magnus-extensions

Extensions packages for magnus

fragler

maraca

R package for the creation of "maraca" plots

PatientSafetyKG

SelfPAD

The official implementation of "Improving Antibody Humanness Prediction using Patent Data".

MCPL

ICML 2024 submission - An image is worth multiple words: learning object level concepts using multi-concepts prompts learning.

qscheck

An R library to perform assertions and decision on input arguments.

molecular-complexity

Python implementation of the molecular complexity metric described by Proudfoot 2017 (http://dx.doi.org/10.1016/j.bmcl.2017.03.008).

Machine-Learning-for-Predicting-Targeted-Protein-Degradation

The code was developed for training diverse ML and DL models to predict PROTACs degradation. Data cleaning for two public datasets, PROTAC-DB and PROTACpedia, are also included. PROTACs are of high interest for all disease areas of AZ and thus predicting their degradation is of general interest.

Jupyter Notebook

OSPred

The OSPred tool offers interactive visualization of clinical trial end point correlations with reference to a large pool of historical NSCLC studies. Its focused capability has the potential to digitally transform and accelerate data-driven decision making as part of the drug development process. OSPred enables data scientists to rapidly visualize, analyze and validate the endpoint (PFS, ORR, OS) correlation hypothesis and to predict HR OS, which potentially could lead to faster and cost-effective NSCLC clinical trials. OSPred - A Digital Health Aid for Rapid Analysis of Early Endpoints (PFS, ORR) In clinical trials that assess novel therapeutic agents in patients with non-small-cell lung cancer (NSCLC), early endpoints (e.g. progression-free survival [PFS] and objective response rate) are often evaluated as indicators of biological drug activity, and are used as surrogate endpoints for overall survival (OS). A data set was compiled to investigate ascertain correlation trends between early endpoints (e.g. odds ratio [OR] for PFS at 6 months) and late endpoints (e.g. hazard ratio [HR] OS). The dataset was curated from multiple source databases, including ClinicalTrials.gov, PubMed and Citeline(TrialTrove). We applied a random-effects method for meta-analysis of prior RCT data to correlate a variety of estimates with the hazard ratio (HR) for OS and PFS. We performed meta-regression analyses across different data-strata, stratified by the mechanism of action as PD1/PDL1, EGFR, VEGFR, DNA and evaluated the correlation of trial-level, treatment effects between early (e.g. PFS) and late (e.g. OS) endpoints in NSCLC oncology trials.

dpp_imp

Improved clinical data imputation via classical and quantum determinantal point processes

survextrap-excesshazards

Demonstration of excess hazard and excess hazard cure models for survival extrapolation

adhce

trim21-bioprotac

Bioinformatics data analyses - Fletcher A. et al., Nature Communications 2023, doi: 10.1038/s41467-023-42546-2

OCT_publication

This repository contains the source code for the image analysis of optical coherence tomography images, as stated in the publication of Volumetric wound healing by machine learning and optical coherence tomography in type 2 diabetes.

lung-tumour-mice-mri

multitask_impute

Supplementary code for 'Deep Learning Imputation for Multi Task Learning'