Costas Maranas

Knowledge-based Database Integration and Development of Pathway Design Tools

This project will be divided into two phases as follows:

(1) Integration of MetRxn with Kbase

The developed MetRxn knowledgebase ( is a unified repository of 8 databases and 112 metabolic models, resulting in 44,784 unique reactions and a million plus unique metabolites. In addition, the database includes 6,211 reaction rules developed using Canonical Labelling for Clique Approximation (CLCA), which leverages prime factorization. In order to make the MetRxn database more accessible to the research community, we will integrate the MetRxn database with KBase ( which is a U.S. department of energy systems biology knowledgebase. This database will serve as a reference for researchers to reconstruct genome-scale model and to investigate metabolic behaviors of complex organisms and microbial communities.

(2) Development of pathway design tool (novoStoic)

The integrated database can then be used to predict novel reactions from enzyme promiscuity and novel pathway combinations accounting for thermodynamic feasibility and protein costs. We will focus on a newly developed novoStoic framework which use standardized MetRxn reaction database and reaction rules to design novel reaction pathways in a mass balanced fashion. While designing novel biotransformations from the substrate to product, we consider design elements such as network size, non-linear pathway topology, mass-conservation, cofactor balance, thermodynamic feasibility and chassis selection.

We expect this project to provide a detailed understanding of pathway design for biofuels or pharmaceuticals in metabolic models. This project will provide multidisciplinary training and education for one undergraduate student as a part of the REU program. The student will perform the simulations for atom mapping, database construction, and subsequently integrate the available visualization tools to build a pathway design community tools.

De novo design of antibodies against zika virus antigen epitopes

Antibody drugs are vital therapeutic workhorses for infectious diseases, cancer, and autoimmune diseases.  However, experimental methods for the generation of therapeutic antibodies such as using immunized mice or directed evolution are not only thwarted by a lack of specificity for specific antigen epitopes but are also time consuming. To this end, we have developed a computational framework called OptMAVEn which can be used to de novo design antibodies against a specific antigen epitope. In this project, we will provide students with a fundamental understanding of our computational suite of programs called Iterative Protein Redesign and Optimization (IPRO), which forms the core computational module for OptMAVEn. We will be working on an antigen epitope of zika virus and de novo design the variable fragment of an antibody. This will give students a brief overview of how a mixed-integer linear program is used to select the optimal combination of V, D and J parts (from the MAPs database) that constitute the antibody variable fragment. Thereafter, we will sort the different antibody designs based on their interaction energies (calculated using CHARMM force field) with the zika antigen epitope. Few top designs will thereafter be passed for MD simulations to check the stability of the antibody-antigen complex. The stable designs will finally be affinity matured using IPRO to improve the binding of the stable designs. Ultimately, these designs will be passed to our experimental collaborators at University of Illinois Chicago, Urbana Champaign, to validate our designs.  

We expect this project to provide an opportunity to learn how to computationally design an antibody against zika virus antigen, that created a stir in recent times due to numerous cases of birth defects, especially neonatal microencephaly, when pregnant women were infected. This project will provide multidisciplinary training and education for one undergraduate student as a part of the REU program. The student will perform the OptMAVEn simulations on a zika epitope and subsequently will take part in discerning the molecular basis of observed interactions between the epitope and various antibody designs.

Faculty Research Links

Contact Information

Manish Kumar, Ph.D.
Assistant Professor of Chemical Engineering
REU Program Coordinator

Esther Gomez, Ph.D.
Assistant Professor of Chemical Engineering
REU Program Coordinator

apply now


The Penn State Department of Chemical Engineering, established in 1948, is recognized as one of the largest and most influential chemical engineering departments in the nation.

The department is built upon the fundamentals of academic integrity, innovation in research, and commitment to the advancement of industry.

Department of Chemical Engineering

119 Greenberg Complex

The Pennsylvania State University

University Park, PA 16802-4400

Phone: 814-865-2574