Enabling Research in Evolutionary Biology
Our focus is on the evolution of hardware and parallel computer architectures as well as on the evolution of molecular sequences.
We understand Bioinformatics as a discipline that develops algorithms, models, and tools that help Biologists to generate new biological insights and knowledge. We try to bridge the gap between the world of systematics and the world of high performance computing.
Due to the increasing descrepancy between the pace of molecular data accumulation and increase in CPU speeds (which is much slower), which we call the "Bio-Gap" we feel that the time has come to establish parallel computing as standard technique in Bioinformatics.
Applications are now open for the 11th summer school on computational molecular evolution organized by Ziheng Yang, Cilia Antoniou, Adam Leache and Alexis Stamatakis. It will be held once again at the EBI in Hinxton, UK.
See the course page for further details
A tool for massively parallel model selection and phylogenetic tree inference on thousands of genes using thousands of cores, based on ModelTest-NG and RAxML-NG. Now published in Bionformatics and available for download here.
The brand new web-server for the completely re-designed version of RAxML, called RAxML-NG is available here.
This is a completely re-designed version of our Evolutionary Placement Algorithm. It is at least 6 times faster than pplacer/EPA and has substantially improved parallel scalability.
It is available for download here
As before EPA-NG user support will be provided via the raxml google group.
This is version 2 of our low level phylogenetic likelihood library.
It now includes site repeats (see paper) to boost computations likelihood computations. It can be downloaded here and has been already integrated into RAxML-NG and Modeltest-NG (see below).
We have completeley re-designed the famous Modeltest software. It's much faster, better, and more stable now.
The all new Modeltest can be downloaded here.
The new, from scratch re-implementation of RAxML is finally avaible at github.
It's faster, more stable, easier to use, and more scalable. application note
libpll is a new library that facilitates phylogenetic analyses. Libpll is a low-level re-design of our previous PLL, that offers unprecedented flexibility when implementing new models, testing hypotheses and/or developing new tools, methods or pipelines.
Apart from the highly-optimized phylogenetic likelihood function, libpll offers the following functionality:
Libpll is highly optimized and can make use of the SSE3/AVX/AVX-2 instruction sets. It is available for download here under the Affero GPL 3 License.
The corresponding documentation can be found here https://github.com/xflouris/libpll/wiki and a large porton of its functionality is presented the following examples.
This tool is a significant improvement of our previous PTP tool for single-locus species delimitation. It is significantlky faster, more accurate, and more scalable with respect to the number of taxa.
The methods are described in this preprint, the code can be downloaded here, and there also is a web-service.
The latest version of RAxML can be found at Alexis github repository.
RAxML questions, help & bug reports: please use the RAxML google group
There is finally a new manual available for RAxML version 8.1.X. You can get it here
We are always looking for student programmers (HiWis) and students interested in doing bachelor/master theses projects with us. If you are interested please send an email to Alexis at Alexandros dot Stamatakis at h hyphen its dot org