What does Moleclue do?

Molecule recommends scientific collaborators who will understand your research and unlock its potential in new ways.



How does molecule work?

We identify researchers with hidden synergies by combining data from every scientific paper in biology, along with biophysical interactions among molecules.  Fundamentally, the structure of the networks we identify provides clues to the molecular biologists who should be working together.



Why should you use it?

Molecule connects you to scientists you would not otherwise meet - but ones who will understand the implications of your topic/molecule and move it in new directions.  


What problem does it solve?

The social and collaborative structure of science is not based on how biology actually works in the cell, which occurs via molecular interactions.  Science is currently organized in terms of reductionist categories, which make it hard to find connections between departments, diseases and topics.  Thus typical attempts to find collaborators (via conferences, friends, linkedin) just keep you in the same old circles.  Even when new dataset arrive in science - which should provide new connections - it takes years for these institutions to react.



What is our solution?

Any topic in molecular biology is meaningfully connected to other topics through molecular interactions that enable life itself.  No human can single-handedly recall these or to understand their full implications for research.  Looking at them as a network can organize workflows in science, as Moleclue connects you to people in your "molecular neighborhood".



What is under the hood?

The entire process of finding collaborators is executed via highly efficient cypher queries in a neo4j database - published details here.  The content in neo4j comes from searching every article on pubmed for molecular topics, identifying the authors of each paper, and assign these to molecular nodes.


Can my company use it?

It is possible to bridge divides in your organization and find innovative solutions with existing resources by using a private version of Moleclue - see contact info.


What features are you planning?

1) Multi-molecule search.

For instance, if an experiment produces hundreds of differentially expressed genes, we can point at an individual who would be most interested in that result, collectively.

2) Author disambiguation

Figuring out unique individuals who wrote each scientific paper is challenging - there can be multiple people working with the same name on similar topics, and most researchers do not have a unique identifier like ORCID.


Where can I learn more?

Questions or feature requests, please email Email to Image Generator


We would like to thank the National Library of Medicine for their assistance in accessing pubmed and Dr. Sara Mostafavi for access to the library of molecular interactions behind Gene Mania, which contribute to search results.