Deep Learning Researcher for Protein Design

About the Role

We are a multidisciplinary research team operating at the intersection of bioinformatics, structural biology, and artificial intelligence. Our primary focus lies in computational protein design, including the creation of novel biomolecules with specified properties for applications in medicine, biotechnology, and basic science. We employ the latest machine learning approaches, particularly generative models and structure prediction methods. We seek experienced individuals ready to engage with leading-edge challenges in the field.

Company Overview

AIRI is a premier Russian institute dedicated to developing fundamental AI models and advancing their practical application in real-world scenarios.

Key Responsibilities

• Develop and implement machine learning and deep learning algorithms targeting de novo protein design, with aims to optimize protein stability, specificity, and functionality.
• Utilize and enhance current architectures such as AlphaFold, RoseTTAFold, ProteinMPNN, and diffusion-based models for protein design tasks.
• Validate in silico findings through collaboration with experimental laboratories.
• Prepare scientific manuscripts and present research outcomes at international conferences.

Mandatory Qualifications

• Strong expertise in machine learning and deep learning algorithms, especially in generative models, graph neural networks, and transformers.
• Experience working with protein structures and sequences, including familiarity with data formats like PDB and mmCIF and libraries such as Biopython, PyRosetta, and MDAnalysis.
• Proficiency in Python programming, alongside knowledge of PyTorch and TensorFlow frameworks.
• Ability to create reproducible analysis pipelines and operate in high-performance computing (HPC) environments.
• Record of publications in reputable scientific journals.
• Experience with Claude Code development environment.

Preferred Skills

• Familiarity with diffusion models applied to protein generation.
• Knowledge of protein structure evaluation methods and free energy calculations using tools like Rosetta, FoldX, and molecular dynamics simulations.
• Experience handling databases such as UniProt, PDB, and AlphaFold DB.
• Participation in projects oriented toward experimental validation.

Working Conditions

• Hybrid working schedule combining office presence and remote work.
• Office located in Moscow City.
• Comprehensive medical insurance (DMS).
• Additional benefits and company discounts.