Open Notebook Science Challenge: Solubilities of Organic Compounds in Organic Solvents

Friesen, Brent; Mancinelli, Jenna; Bohinski, Tim; Mirza, Khalid B.; Bulger, David; Moritz, Marshall; Federici, Matthew; Rein, Daniel T.; Tchakounte, Cedric; Bradley, Jean‐Claude; Truong, Hai Bang; Neylon, Cameron; Guha, Rajarshi; Williams, Antony J.; Hooker, Bill; Hale, Jenny; Lang, Andrew

doi:10.1038/npre.2010.4243.1

Cited by 3 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, they are commonly used in chemistry applications with SMILES sequences. [58,59] In our regression model, we use SELFIES because it matches the representation used in MACCS. However, using SELFIES over SMILES does not necessarily translate to better supervised learning performance.…”

Section: Small Molecule Solubility Predictionmentioning

confidence: 99%

Model agnostic generation of counterfactual explanations for molecules

Wellawatte

Seshadri

White

2021

Preprint

View full text Add to dashboard Cite

An outstanding challenge in deep learning is its lack of interpretability. The inability of both designers and users of neural networks to explain why a prediction is made is a major drawback. This not only dissuades chemists from using deep learning predictions, but also has led to neural networks learning spurious correlations that are difficult to notice. Counterfactuals are a category of prediction explanations that provide a rationale behind the prediction value. Counterfactuals have satisfying properties like being in the same space as the features and require no deep learning expertise to understand. Counterfactuals are generated via optimization, which makes them complex to use when the features are chemical structures. Here we present an algorithm built on the "Superfast Traversal, Optimization, Novelty, Exploration and Discovery" method that is efficient and model-agnostic for generating counterfactuals. We show this method is applicable in random forest models, sequence models, and graph neural networks in both classification and regression. Our method requires no additional training, data, or model gradients.

show abstract

Section: Small Molecule Solubility Predictionmentioning

confidence: 99%

Model agnostic generation of counterfactual explanations for molecules

Wellawatte

Seshadri

White

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Thus, they are commonly used in chemistry applications with SMILES sequences. [59,60] In our regression model, we use SELFIES because it matches the representation used in MACCS. However, using SELFIES over SMILES does not necessarily translate to better supervised learning performance.…”

Section: Small Molecule Solubility Predictionmentioning

confidence: 99%

Model agnostic generation of counterfactual explanations for molecules

Wellawatte

Seshadri

White

2021

Preprint

View full text Add to dashboard Cite

An outstanding challenge in deep learning in chemistry is its lack of interpretability. The inability of explaining why a neural network makes a prediction is a major barrier to deployment of AI models. This not only dissuades chemists from using deep learning predictions, but also has led to neural networks learning spurious correlations that are difficult to notice. Counterfactuals are a category of explanations that provide a rationale behind a model prediction with satisfying properties like providing chemical structure insights. Yet, counterfactuals are have been previously limited to specific model architectures or required reinforcement learning as a separate process. In this work, we show a universal model-agnostic approach that can explain any black-box model prediction. We demonstrate this method on random forest models, sequence models, and graph neural networks in both classification and regression.

show abstract

“…Cheminformaticians will recognize this data structure as an edge from graph theory. This structure allows us to represent facts like "vanillin dissolves in methyl alcohol" [ 7 ]. RDF uses Uniform Resource Identifiers (URIs) to identify things.…”

Section: Conceptsmentioning

confidence: 99%