AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

CogMol: Target-Specific and Selective Drug Design for COVID-19 Using Deep Generative Models

Vijil Chenthamarakshan, Payel Das, Samuel Hoffman, Hendrik Strobelt, Inkit Padhi, Kar Wai Lim, Benjamin Hoover, Matteo Manica, Jannis Born, Teodoro Laino, Aleksandra Mojsilovic

NeurIPS 2020 · Probabilistic and scientific modeling › Molecules and proteins · approach: Structure-based 3D molecule generation · 82 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use CLaSS guided sampling instead of random latent-space samplingCogMol · SARS-CoV-2 targets, affinity, QED, and selectivity criteriaaccepted fraction: NSP9: 0.007 -> 0.069; RBD: 0.009 -> 0.09; MPro: 0.011 -> 0.104Improved
Add QED control to affinity control (ablation)CogMol CLaSS · NSP9accepted fraction: 0.567 -> 0.45Hurt
Add QED control to affinity control (ablation)CogMol CLaSS · RBDaccepted fraction: 0.546 -> 0.429Hurt
Add QED control to affinity control (ablation)CogMol CLaSS · MProaccepted fraction: 0.603 -> 0.472Hurt
Add selectivity control to affinity and QED controls (ablation)CogMol CLaSS · NSP9accepted fraction: 0.45 -> 0.069Hurt
Add selectivity control to affinity and QED controls (ablation)CogMol CLaSS · RBDaccepted fraction: 0.429 -> 0.09Hurt
Add selectivity control to affinity and QED controls (ablation)CogMol CLaSS · MProaccepted fraction: 0.472 -> 0.104Hurt
Use SMILES representation instead of VAE latent representation for affinity prediction (ablation)Binding affinity predictor · BindingDB test databinding affinity RMSE: 1.2820 -> 0.8426Improved
Use pretrained protein embeddings rather than LSTM protein embeddings (ablation)Binding affinity predictor · BindingDB test databinding affinity RMSE: 1.0104 -> 0.8426Improved
Adaptively train the VAE on BindingDB rather than ZINC (ablation)Molecular VAE · Molecular generationvalid molecule fraction: 0.9553 -> 0.904Hurt
Use MT-DNN rather than random forest for toxicity prediction (ablation)Toxicity predictor · Tox21 and ClinTox averageF1: 0.36 -> 0.45Improved
Use MT-DNN rather than random forest for toxicity prediction (ablation)Toxicity predictor · Tox21 and ClinTox averageROC AUC: 0.81 -> 0.78Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy