Personal tools
You are here: Home People Loewe, Laurence

Homepage of Laurence Loewe

A short overview of why I am doing what I am doing, my CV, research interests, publications and teaching.

 

We are transitioning to a new website,
which you can find here
. Until the new site is complete, you can still find useful content on this site. Thank you for your patience.

 

Photo2006-150px

 


 

 

 


 


 

 

Current Address:

Laurence Loewe
Assistant Professor 
Laboratory of Genetics and
Wisconsin Institute for Discovery
University of Wisconsin-Madison

Mailing address:
330 North Orchard Street,
Madison, WI, 53715, USA

Tel: +1(608)-316-4324
Email: Laurence.Loewe@evolutionary-research.net

If urgent, please email me at l o e w e
at w i s c dot e d u.











Modeling is at the heart of understanding biology - consciously or subconsciously.

My research career before Madison is easily summarized as years of building models in population genetics and years of building models in molecular systems biology. I learned what it takes to build a model that is relevant, reasonably reliable and usable. Many critical details in this modeling process are repetitive, cumbersome and error prone if performed by humans. Computers could do this much better - if there was a way to tell them what to do. I have seen enough to convince me that this is possible in general, but I have yet to see a system that can do it as needed. Thus, much work in computational biology and bioinformatics remains marred by what I have come to call 'silicon digging' and 'data shoveling': the use of manual labor to substitute for the lack of appropriate algorithms.

Working in theoretical population genetics and molecular systems biology made me appreciate the enormous integrative power of evolution as a concept. This inspired me to help build a bridge between current systems biology and evolutionary biology. This led me to propose a framework for evolutionary systems biology, or 'EvoSysBio' in short. This new field critically depends on the quality of the computational models it constructs. They have to be rigorous, reliable and realistic - a tall order in a world where the state of the art keeps progressing, new data keeps arriving, and software is rife with version conflicts that make it difficult to use today's code with tomorrows versions. For over 10 years I have been searching for tools that support all it takes to keep the focus on biology, without the need to laboriously work around some serious deficiencies. I got tired of trying out tools.

After checking my chances, I decided in 2014 that my best bet of effectively getting EvoSysBio off the ground is to write a new programming language. Enter Evolvix.

Evolvix is the first general programming language designed by biologists for biologists

... at least to my knowledge. While not limited to biology, biologists are its prime intended audience. Hence, when deciding how to do things in Evolvix, I strive to translate new features into words understood by those biologists, who have otherwise given up on math and computing.

Why?

If EvoSysBio can't leverage their expertise in describing biological reality, then the range of scientific problems that EvoSysBio will be able to address will remain severely limited.

By now many of the broad design questions in Evolvix are conceptually resolved, but many details remain to be worked out. These will be addressed one after another, as needed and possible.

As of April 2015, my research group is working to implement a prototype for the general core, while I am catching up with writing up scientific papers, updating websites, writing up Evolvix requirements, specifications and designs, etc...

If you are an able C++11, JavaScript or Python developer or are excited about building compilers and are looking for the opportunity to contribute towards an (eventually) open-source software project in a fast-paced, highly dynamic research environment, then please check the jobs page and contact me. We are always looking for outstanding programmers who can follow industry best-practices.

We will also be looking for an editor to support the work.

If you are working in computational biology of any kind or are interested in otherwise contributing towards designing or implementing Evolvix, please let me know by emailing my work account (at wisc.edu). My goal is to put Evolvix on a path, where it becomes easy to integrate new tools in a way that allows them to become stable and reliable. We are currently refining various approaches for how to do this. I can't promise to include all submissions; in fact I will not, for many reasons. However, I aim to review and consider all submissions. If they are good, I want them to shape my thinking and I aim to allocate for them a place in the Evolvix concept and syntax space that will allow someone to implement them in a way that integrates well.

The overall goal: make it easy for researchers to put the "properly quantified" into the next sentence.

 

Nothing in biology makes sense except when properly quantified in the light of evolution.

Evolutionary processes are at the heart of many problems that we face in our world today, ranging from antibiotics resistance evolution to species extinction. Addressing such problems requires models of the underlying causes. I aim to improve the quality of these models by quantifying evolution with increasing precision.

To this end I estimate the strength of selection in various systems, using different approaches, including the analysis of DNA sequences by population genetics methods. I also develop a new approach that builds on existing quantitative models from current systems biology and links them to potential fitness correlates to help estimate distributions of mutational effects in silico. This is an important part of EvoSysBio and combines the strengths of evolutionary genetics and systems biology.

All this is facilitated by Evolvix (see above), which helps me to split off the computational work from the biology work in my computational biology. I can then bundle it into programming times, where I can deal with it more efficiently, and where it does not interrupt my biological train of thought. This approach helps me to improve models for systems biology, genome evolution, antibiotics resistance evolution and species extinction - without having to re-implement the same features over and over again for different models.

If you want to join us in the Evolutionary Systems Biology Group at the Wisconsin Institutes for Discovery, please check out the jobs page.

While I am re-organizing my website, you can use my existing research interests pagepublications page, and teaching page.

Other webpages associated with me:

Other websites associated with me:

  • EvoSysBio-course.discovery.wisc.edu - Website for the 3 credit EvoSysBio Course at UW-Madison that I have been teaching each Fall since 2013. Registration open for Fall 2015
  • Evolvix.org - Homepage of the Evolvix model description and programming language for biology.
  • EvolutionarySystemsBiology.org - Pointer to some EvoSysBio activities, labs and resources.
  • Evolution-at-home.org - the first global computing project for evolutionary biology that I started in 2001 and that is in the process of being completely redesigned using Evolvix.
  • QuantBio.wisc.edu - A brief overview of some activities in quantitative biology at the University of Wisconsin-Madison.

 

 

Short CV

 

 

Document Actions