CCB Brown Bag Seminar via Zoom

Date & Time

Speaker: Vladimir Gligorijevic, Ph.D., Research Scientist, Systems Biology

Topic: Structure-based region-specific protein function prediction and conditional protein design using deep learning

With the maturing of de novo structure prediction methods and the rise of deep learning techniques, it now becomes possible to generate high-throughput structure and function predictions for many unannotated proteins.

We will first introduce deepFRI (deep functional residue identification), our recently proposed geometric deep learning method based on Graph Convolutional Networks (GCNs). As opposed to CNNs, that have been the state-of-the-art in predicting protein function from sequence, we show that GCNs are better at extracting features from proteins and predicting their functions by taking into account the graph-based structure of their amino acid residues represented by contact maps. deepFRI not only improves predictions of Gene Ontology (GO) terms from protein sequences and predicted 3D structures, but also brings residue-level saliency mapping. The mapping provides insight into putative functional sites allowing for biological interpretation, hypothesis generation or the design of targeted validation experiments.

In the second half of the talk, I will discuss our new generative models for graph-based conditional protein design. I will introduce a novel architecture that combines graph denoising autoencoders (GDAE) with an external function analyser. GDAE, composed of graph convolutional encoder with a self-attention decoder, learns the low-dimensional manifold of protein structures. GDAE is used for sampling novel protein structures consistent with the local manifold structure that are further optimized for desired functions using a pre-trained deepFRI architecture in combination with Rosetta energy terms as the function analyser.

Zoom Info:

Join from PC, Mac, Linux, iOS or Android:

Or Telephone:
Dial(for higher quality, dial a number based on your current location):
US: +1 646 558 8656 or +1 669 900 6833
Meeting ID: 752 878 068
International numbers available:

Or an H.323 room system: (US East)
Meeting ID: 752 878 068

Advancing Research in Basic Science and MathematicsSubscribe to Flatiron Institute announcements and other foundation updates