Life Sciences App Catalog

Drug Discovery

This Dash app demonstrates a simple interface for comparing molecule features and distributions for drug discovery. Drug candidates can be highlighted and added to a table based on their molecular weight, structure, chirality, and pKa.

This app's source code is available in Python and R.
View the App.

Deep Learning Compound Screening

This Dash app, developed by Recursion Pharma, is an interface to a deep learning model of cellular morphology, demonstrating a scalable “phenomics” platform. It presents the ability to identify dose-dependent, high-dimensional relationships among and between immunomodulators, toxins, pathogens, genetic perturbations, and small and large molecules at scale.

Access this Dash app and get the Python code. 

ML Assisted CRISPR

This Dash app, published in Nature by MIT’s Gifford Lab, integrates known biological mechanisms of DNA repair with machine learning to achieve strong accuracy in predicting CRISPR outcomes with the inDelphi prediction model.

Access this Dash app and get the Python code.

Pharmacokinetics Calculator

This Dash app demonstrates a simple graphical user interface (GUI) for a pharmacokinetics calculator. This GUI enables parametrized, non-compartmental analysis of the effects of the body on a drug.

Access this Dash app and get the Python code. 

COVID Diagnostic Accuracy

Developed by In-Q-Tel's B.Next lab for biotechnology and U.S. national security, this Dash app allows users to explore COVID diagnostic test performance expectations based on reported clinical trial data. The Dash app allows you to select performance data from 269 studies covering hundreds of diagnostic tests for coronavirus to examine expected positive and negative predictive values (PPV and NPV respectively) across a disease prevalence range.

Access this Dash app and get the Python code.

US Opioid Epidemic

This Dash app explores the effect of the opioid epidemic in the United States by county with dynamic cross-filtering across components. The Dash app plots poison-induced mortality rates sourced from the CDC Wonder public health information system, with cause of deaths classified by the International Classification of Diseases, Tenth Revision (ICD-10).

Access this Dash app and get the Python code. 

COVID SEIR Modeling

This Dash app was developed by the MIT Operations Research Center. The project recognizes that, “A critical tool for COVID-19 planning is charting out the progression of the pandemic across the United States and the world. We’ve developed a new epidemiological model called DELPHI, which forecasts infections, hospitalizations, and deaths. You can think of our model as a standard SEIR model with additional features specific to the COVID-19 pandemic, like under-detection and differentiated government intervention.”

Access this Dash app and get the Python code. 

COVID Chest X-Rays

This Dash app allows users to annotate medical image datasets of COVID-19 patients in order to extract a lung occlusion (and train machine learning models).

Access this Dash app and get the Python code. 

Heart Disease Prediction   

This Dash app uses IBM’s explainable AI (“xAI”) Python library to predict heart disease. The app displays the training and test accuracy of the model and includes an interface to filter through and visualize the generalized additive model (GAM) for the feature-set.

View this community forum post to learn more.

Access this Dash app and get the Python code.

Transcriptome Sequencing QA

This Dash app was developed by the Ontario Institute for Cancer Research (OICR). OICR has developed a quality assurance (QA) system in Dash that is currently in production for OICR transcriptome sequencing assays. The Dash QA system shows human and virus genomes alongside each other and key metrics for evaluating whether sequencing experiments have succeeded.

Access this Dash app and get the Python code. 

Phylogenetic Tree Reconciliation  

This Dash app was published in Oxford Bioinformatics by the COBIUS lab at University of Sherbrooke. It visualizes double reconciliations between phylogenetic trees at three levels: transcript, gene and species. The app extends the RecPhyloXML model—developed for gene–species tree reconciliation—to represent joint transcript–gene and gene–species tree reconciliations.

Access this Dash app and get the Python code.