Michael Craig - Machine Learning on molecular data

PyData London Meetup #46
Tuesday, July 3, 2018

At GTN we are combining ideas from quantum physics and chemistry with machine learning to aid the process of discovering new medicines. In this talk I will discuss the challenges of applying machine learning to molecular datasets. Issues of data representation are starkly different then for, say, image or text based data, and I will describe various ways to represent molecules, starting with simple representations such as SMILES strings, chemical fingerprints, and going on to more advanced graph based, and quantum mechanical representations. In going beyond text or matrix based representations to graphs, standard convolutional or RNN networks are no longer appropriate, and recently developed architectures specifically tailored for graph data need to be utilised. I will describe recent advances in so called “graph convolutional networks” that have generated best in class results on chemistry datasets, and will demo how one can curate molecular datasets from public sources, most prominently ChEMBL, and run advanced ML algorithms on them using publicly available python libraries such as RDKIT and deepchem.

Sponsored & Hosted by Man AHL

****

www.pydata.org

PyData is an educational program of NumFOCUS, a 501(c)3 non-profit organization in the United States. PyData provides a forum for the international community of users and developers of data analysis tools to share ideas and learn from each other. The global PyData network promotes discussion of best practices, new approaches, and emerging technologies for data management, processing, analytics, and visualization. PyData communities approach data science using many languages, including (but not limited to) Python, Julia, and R.

PyData conferences aim to be accessible and community-driven, with novice to advanced level presentations. PyData tutorials and talks bring attendees the latest project features along with cutting-edge use cases.

Видео Michael Craig - Machine Learning on molecular data канала PyData