Predicting chemical properties from text expression of molecules

[0sophy1]

Abstract

Text representation - SMILE - can contain simple molecular structure information together with its atomic composition. Let's try to find the chemical properties of molecules by learning SMILE.

Description

Predicting chemical properties - solubility, acidity, etc - is one of the important issues on chemical research including drug discovery. There are several trials and result in getting meaningful prediction results on classical NN: https://arxiv.org/abs/1712.02034 https://arxiv.org/pdf/1811.08283.pdf https://molecularexploration.medium.com/smiles-toxicity-prediction-e75b27863da5

Getting ideas from classical ML, let's do this with quantum computers. If you're already familiar with NLP or QNLP, consider whether you can use entanglement between words (or subtext groups in SMILE) to make better guesses. https://github.com/oxford-quantum-group/discopy/blob/main/docs/notebooks/qnlp-tutorial.ipynb

Members

• @githubhandle
• @githubhandle - Slack: @slackhandle email: example@example.com
• Qiskit Coach: @0sophy1

Deliverable

GitHub repo

[0sophy1]

mentor not available.. cheer Sophy