A multilayer network approach to quantifying biologically-derived systematic risk in biomedical finance

Sharply rising disease prevalence and associated healthcare costs are placing an increasingly significant economic burden on society. Biomedical research and industry have struggled to adequately address this challenge, as evidenced by the stagnation and even decline of new therapeutics development success rates. Recent work in the MIT Laboratory for Financial Engineering has explored the potential of using financial engineering in the form of biomedical "megafunds" to help tackle this problem. New methods will be needed to better assess systematic financial risks for these therapeutic project portfolios. This primarily methodological thesis seeks to explore the opportunity to leverage multilayer network models as tools to help measure this risk, specifically the biologically-derived component of risk resulting from project correlations generated through the underlying biological networks. Historical examples of coupling between drug development projects are used to motivate a framework in which project correlations emerge from a combination of indication and target similarity. This framework motivates the construction of a multilayer network model, drawing upon multiple systems biology databases for its construction and using a sample of FDA orphan designations as a representative project set. Using shortest path distance and Random Walk with Restart (RWR) relevance, indication and target similarity between projects are quantitatively evaluated. Comparing average sales correlations to the log of average RWR relevance for classes of compounds reveals notable relationships between correlation and network similarity. This relationship is shown to be stronger for the case of disease relevance (R2 = 0.99) than for target relevance (R2 = 0.93). A potential approach is finally described for integrating biological network similarity with financial models useful for portfolio analysis, and implications on portfolio selection are discussed through synthetic construction of hypothetical orphan drug portfolios..

Department dc:contributor.department

Massachusetts Institute of Technology. Engineering and Management Program

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2016

Author dc:creator

• Walz, Andrew R

Advisor dc:contributor.advisor

• Olivier de Weck and Andrew Lo.

• Engineering and Management Program.
• System Design and Management Program.
• Engineering Systems Division.

Statement dc:rights

• MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/107356

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/107356