English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We present a novel multimodal language model approach for predictingmolecular properties by combining chemical language representation withphysicochemical features. Our approach, MULTIMODAL-MOLFORMER, utilizes a causalmultistage feature selection method that identifies physicochemical featuresbased on their direct causal effect on a specific target property. These causalfeatures are then integrated with the vector space generated by molecularembeddings from MOLFORMER. In particular, we employ Mordred descriptors asphysicochemical features and identify the Markov blanket of the targetproperty, which theoretically contains the most relevant features for accurateprediction. Our results demonstrate a superior performance of our proposedapproach compared to existing state-of-the-art algorithms, including thechemical language-based MOLFORMER and graph neural networks, in predictingcomplex tasks such as biodegradability and PFAS toxicity estimation. Moreover,we demonstrate the effectiveness of our feature selection method in reducingthe dimensionality of the Mordred feature space while maintaining or improvingthe model's performance. Our approach opens up promising avenues for futureresearch in molecular property prediction by harnessing the synergisticpotential of both chemical language and physicochemical features, leading toenhanced performance and advancements in the field.

Beyond Chemical Language: A Multimodal Approach to Enhance Molecular Property Prediction