Mots-Clés

Synthetic Biology, Retrosynthesis, Chemistry, Computational Chemistry, Artificial Intelligence, Bioinformatics

Description

1. Context

We are seeking a Data Scientist to contribute to our on-going effort in developing retrosynthesis methods (cf. https://www.jfaulon.com). Retrosynthesis is the process of “deconstructing” a target molecule into readily available starting materials (i.e., molecules provided by chemical vendors or naturally produced by living organisms). Studied in chemistry since the 1970s, then in biology since the 2010s, retrosynthesis is now a subject of research for the AI/Computer Science community (cf. https://arxiv.org/abs/2301.05864).

The candidate will take part in research projects funded by the French funding research agency (ANR) and the Ile-de-France region at the Micalis Institute. Micalis (INRAE & AgroParisTech, https://www.micalis.fr) is a research unit of more than 350 researchers developing multidisciplinary approaches and promoting the development of synthetic biology applications for health and biotechnology. Within Micalis, the recruiting research team (~20 staff spread into both wet and dry labs) specializes in developing computer aided design tools for biotechnology including retrosynthesis methods. The candidate will closely collaborate with software developers and interact with Research Scientists, Engineers and PhD fellows.

2. Mission(s)

We are seeking a candidate to work on novel AI-driven retrosynthesis methods to be applied to synthesis planning in chemistry, retro-biosynthesis in biotechnology and mixed chemical/biological retrosynthesis. The candidate’s tasks include:

- Develop and benchmark deterministic and AI methods for template-based and template-free one-step retrosynthesis.

- Embed on-step retrosynthesis within MCTS or AI* search algorithms.

- Develop methods to perform retrosynthesis with both chemical reactions and biological (enzymatic) reactions. 

- Develop methods to perform retro-biosynthesis with consortia of organisms. 

3. Profile

- PhD degree in Data science/AI or Computer Science, Cheminformatics, Bioinformatics.

- Prior knowledge with Transformer and Graph Neural Network architectures along with machine learning libraries (TensorFlow, PyTorch).

- Strong programming skills (Python, C++).

- Good knowledge of Unix-like OSs, testing practices and Git versioning are required.

- Knowledge in continuous integration paradigms is a plus.

4. Contract

- Fixed-term contract of 12 months (can be renewed twice), in accordance with the French legislation.