Python Tools for Computational Molecular Biology
Python Tools for Computational Molecular Biology
For the past several years, Biopython-related GSoC projects have been successfully run under the mentorship of the Open Bioinformatics Foundation (OBF) and the National Evolutionary Synthesis Center (NESCent).
Please read those organizations’ GSoC pages and the main Google Summer of Code page for more details about the program.
See the Open Bioinformatics Foundation (OBF) GSoC wiki page and OBF GSoC page as our usual mentoring organization.
Any project ideas for Biopython can be posted at http://obf.github.io/GSoC/ideas/ (update https://github.com/OBF/GSoC/blob/gh-pages/00_ideas.md via a GitHub pull request).
We encourage potential students and mentors to join the BioPython mailing lists and actively participate in developing these project ideas, either by submitting their own ideas or contributing to improving existing ones.
Usually, each BioPython proposal has one or more mentors assigned to it. Nevertheless, we encourage potential students/mentors to contact the mailing lists with their own ideas for proposals. There is therefore not a set list of ‘available’ mentors, since it highly depends on which projects are proposed every year.
Past mentors include:
Student Evan Parker (blog)
Rationale Bio.SeqIO’s indexing offers parsing on demand access to any sequence in a large file (or collection of files on disk) as a SeqRecord object. This works well when you have many small to medium sized sequences/genomes. However, this is not ideal for large genomes or chromosomes where only a sub-region may be needed. A lazy-loading parser would delay reading the record until requested. For example, if region record[3000:4000] is requested, then only those 1000 bases need to be loaded from disk into memory, plus any features in that region. This is how Biopython’s BioSQL interface works. Tools like tabix and samtools have demonstrated efficient co-ordinate indexing which could be useful here.
Aside from being used via an index for random access, lazy-loading parsers could be used when iterating over a file as well. This can potentially offer speed ups for tasks where only a fraction of the data is used. For example, if calculating the GC content of a collection of genomes from GenBank, using Bio.SeqIO.parse(…) would currently needlessly load and parse all the annotation and features. A lazy-parser would only parse the sequence information.
Approach & Goals Useful features include:
Difficulty and needed skills Hard. Familiarity with the Biopython’s existing sequence parsing essential. Understanding of indexing large files will be vital.
Possible Mentors Wibowo Arindrarto, Peter Cock, others welcome
This year the Open Bioinformatics Foundation was not accepted on the very competitive GSoC programme. Biopython instead participated under NEScent.
Student Zheng Ruan (blog)
Rationale A codon alignment is an alignment of nucleotide sequences in which the trinucleotides correspond directly to amino acids in the translated protein product. This carries important information which can be used for several analysis, notably analysis of selection pressures. This project extends Biopython to support this data type and these analyses.
Approach & Goals Useful features include:
Difficulty and needed skills Medium, depending on ambition. Familiarity with the Biopython’s existing alignment classes and functions, or equivalents in BioPerl, BioJava or BioRuby (e.g.), will be helpful. Understanding of the practical uses of codon alignments, or at least a basic understanding of molecular biology, is important. Some basic math is involved, essentially reading a few equations and converting them to code. One useful book to have on hand is Computational Molecular Evolution by Ziheng Yang.
Mentors Eric Talevich, Peter Cock
Student Yanbo Ye (blog)
Rationale While the Phylo module in Biopython supports I/O and basic tree operations, there are some important components that remain to be implemented to better support phylogenetic workflows.
Approach & Goals This “idea” is intentially left open-ended – some potentially useful features are:
Difficulty and needed skills Medium-to-advanced programming skill in Python – it’s important for these implementations to be reasonably efficient, though we don’t aim to compete with the fastest stand-alone implementations of these algorithms. Knowledge of phylogenetic methods is critical; for reference, you might like to have a copy of Joe Felsenstein’s Inferring Phylogenies. Tree visualizations are done with matplotlib.
Mentors Mark Holder, Jeet Sukumaran, Eric Talevich
Student Wibowo Arindrarto (blog)
Rationale Biopython has general APIs for parsing and writing assorted sequence file formats (SeqIO), multiple sequence alignments (AlignIO), phylogenetic trees (Phylo) and motifs (Bio.Motif). An obvious omission is something equivalent to BioPerl’s SearchIO. The goal of this proposal is to develop an easy-to-use Python interface in the same style as SeqIO, AlignIO, etc but for pairwise search results. This would aim to cover EMBOSS muscle & water, BLAST XML, BLAST tabular, HMMER, Bill Pearson’s FASTA alignments, and so on.
Approach Much of the low level parsing code to handle these file formats already exists in Biopython, and much as the SeqIO and AlignIO modules are linked and share code, similar links apply to the proposed SearchIO module when using pairwise alignment file formats. However, SearchIO will also support pairwise search results where the pairwise sequence alignment itself is not available (e.g. the default BLAST tabular output). A crucial aspect of this work will be to design a pairwise-search-result object heirachy that reflects this, probably with a subclass inheriting from both the pairwise-search-result and the existing MultipleSequenceAlignment object. Beyond the initial challenge of an iterator based parsing and writing framework, random access akin to the Bio.SeqIO.index and index_db functionality would be most desirable for working with large datasets.
Challenges The project will cover a range of important file formats from major Bioinformatics tools, thus will require familiarity with running these tools, and understanding their output and its meaning. Inter-converting file formats is part of this.
Difficulty and needed skills Medium/Hard depending on how many objectives are attempted. The student needs to be fluent in Python and have knowledge of the BioPython codebase. Experience with all of the command line tools listed would be clear advantages, as would first hand experience using BioPerl’s SearchIO. You will also need to know or learn the git version control system.
Mentors Peter Cock
Student Lenna Peterson (blog)
Rationale Computational analysis of genomic variation requires the ability to reliably communicate and manipulate variants. The goal of this project is to provide facilities within BioPython to represent sequence variation objects, convert them to and from common human and file representations, and provide common manipulations on them.
Approach & Goals
Difficulty and needed skills Easy-to-Medium depending on how many objectives are attempted. The student will need have skills in most or all of: basic molecular biology (genomes, transcripts, proteins), genomic variation, Python, BioPython, Perl, BioPerl, NCBI Eutilities and/or Ensembl API. Experience with computer grammars is highly desirable. You will also need to know or learn the git version control system.
Mentors Reece Hart
Student Mikael Trellet
Rationale Analysis of protein-protein complexes interfaces at a residue level yields significant information on the overall binding process. Such information can be broadly used for example in binding affinity studies, interface design, and enzymology. To tap into it, there is a need for tools that systematically and automatically analyze protein structures, or that provide means to this end. Protorop (http://www.bioinformatics.sussex.ac.uk/protorp/) is an example of such a tool and the elevated number of citations the server has had since its publication acknowledge its importance. However, being a webserver, Protorop is not suited for large-scale analysis and it leaves the community dependent on its maintainers to keep the service available. On the other hand, Biopython’s structural biology module, Bio.PDB, provides the ideal parsing machinery and programmatic structures for the development of an offline, open-source library for interface analysis. Such a library could be easily used in large-scale analysis of protein-protein interfaces, for example in the CAPRI experiment evaluation or in benchmark statistics. It would be also reasonable, if time permits, to extend this module to deal with protein-DNA or protein-RNA complexes, as Biopython supports nucleic acids already.
Approach & Goals
Difficulty and needed skills Easy/Medium. Working knowledge of the Bio.PDB module of BioPython. Knowledge of structural biology in general and associated file formats (PDB).
Mentors João Rodrigues
Student Michele Silva
Rationale Discovering the structure of biomolecules is one of the biggest problems in biology. Given an amino acid or base sequence, what is the three dimensional structure? One approach to biomolecular structure prediction is the construction of probabilistic models. A Bayesian network is a probabilistic model composed of a set of variables and their joint probability distribution, represented as a directed acyclic graph. A dynamic Bayesian network is a Bayesian network that represents sequences of variables. These sequences can be time-series or sequences of symbols, such as protein sequences. Directional statistics is concerned mainly with observations which are unit vectors in the plane or in three-dimensional space. The sample space is typically a circle or a sphere. There must be special directional methods which take into account the structure of the sample spaces. The union of graphical models and directional statistics allows the development of probabilistic models of biomolecular structures. Through the use of dynamic Bayesian networks with directional output it becomes possible to construct a joint probability distribution over sequence and structure. Biomolecular structures can be represented in a geometrically natural, continuous space. Mocapy++ is an open source toolkit for inference and learning using dynamic Bayesian networks that provides support for directional statistics. Mocapy++ is excellent for constructing probabilistic models of biomolecular structures; it has been used to develop models of protein and RNA structure in atomic detail. Mocapy++ is used in several high-impact publications, and will form the core of the molecular modeling package Phaistos, which will be released soon. The goal of this project is to develop a highly useful Python interface to Mocapy++, and to integrate that interface with the Biopython project. Through the Bio.PDB module, Biopython provides excellent functionality for data mining biomolecular structure databases. Integrating Mocapy++ and Biopython will allow training a probabilistic model using data extracted from a database. Integrating Mocapy++ with Biopython will create a powerful toolkit for researchers to quickly implement and test new ideas, try a variety of approaches and refine their methods. It will provide strong support for the field of biomolecular structure prediction, design, and simulation.
Approach & Goals Mocapy++ is a machine learning toolkit for training and using Bayesian networks. It has been used to develop probabilistic models of biomolecular structures. The goal of this project is to develop a Python interface to Mocapy++ and integrate it with Biopython. This will allow the training of a probabilistic model using data extracted from a database. The integration of Mocapy++ with Biopython will provide a strong support for the field of protein structure prediction, design and simulation.
Mentors Eric Talevich
Student Justinas V. Daugmaudis
Rationale BioPython is a very popular library in Bioinformatics and Computational Biology. Mocapy++ is a machine learning toolkit for parameter learning and inference in dynamic Bayesian networks (DBNs), which encode probabilistic relationships among random variables in a domain. Mocapy++ is freely available under the GNU General Public Licence (GPL) from SourceForge. The library supports a wide spectrum of DBN architectures and probability distributions, including distributions from directional statistics. Notably, Kent distribution on the sphere and the bivariate von Mises distribution on the torus, which have proven to be useful in formulating probabilistic models of protein and RNA structure. Such a highly useful and powerful library, which has been used in such projects as TorusDBN, Basilisk, FB5HMM with great success, is the result of the long-term effort. The original Mocapy implementation dates back to 2004, and since then the library has been rewritten in C++. However, C++ is a statically typed and compiled programming language, which does not facilitate rapid prototyping. As a result, currently Mocapy++ has no provisions for dynamic loading of custom node types, and a mechanism to plug-in new node types that would not require to modify and recompile the library is of interest. Such a plug-in interface would assist rapid prototyping by allowing to quickly implement and test new probability distributions, which, in turn, could substantially reduce development time and effort; the user would be empowered to extend Mocapy++ without modifications and subsequent recompilations. Recognizing this need, the project (herein referred as MocapyEXT), with the aim to improve the current Mocapy++ node type extension mechanism, has been proposed by T. Hamelryck.
Approach & Goals The MocapyEXT project is largely an engineering effort to bring a transparent Python plug-in interface to Mocapy++, where built-in and dynamically loaded node types could be used in a uniform manner. Also, externally implemented and dynamically loaded nodes could be modified by a user and these changes will not necessitate the recompilation of the client program, nor the accompanying Mocapy++ library. This will facilitate rapid prototyping, ease the adaptation of currently existing code, and improve the software interoperability whilst introducing minimal changes to the existing Mocapy++ interface, thus facilitating a smooth acceptance of the changes introduced by MocapyEXT.
Mentors Eric Talevich
Student João Rodrigues
Rationale Biopython is a very popular library in Bioinformatics and Computational Biology. Its Bio.PDB module, originally developed by Thomas Hamelryck, is a simple yet powerful tool for structural biologists. Although it provides a reliable PDB parser feature and it allows several calculations (Neighbour Search, RMS) to be made on macromolecules, it still lacks a number of features that are part of a researcher’s daily routine. Probing for disulphide bridges in a structure and adding polar hydrogen atoms accordingly are two examples that can be incorporated in Bio.PDB, given the module’s clever structure and good overall organisation. Cosmetic operations such as chain removal and residue renaming – to account for the different existing nomenclatures – and renumbering would also be greatly appreciated by the community. Another aspect that can be improved for Bio.PDB is a smooth integration/interaction layer for heavy-weights in macromolecule simulation such as MODELLER, GROMACS, AutoDock, HADDOCK. It could be argued that the easiest solution would be to code hooks to these packages’ functions and routines. However, projects such as the recently developed edPDB or the more complete Biskit library render, in my opinion, such interfacing efforts redundant. Instead, I believe it to be more advantageous to include these software’ input/output formats in Biopython’s SeqIO and AlignIO modules. This, together with the creation of interfaces for model validation/structure checking services/software would allow Biopython to be used as a pre- and post-simulation tool. Eventually, it would pave the way for its inclusion in pipelines and workflows for structure modelling, molecular dynamics, and docking simulations.
Mentors Eric Talevich
Diana Jaunzeikare
This was the first year Biopython took part in GSoC, and we did so under the banner of NESCent’s GSoC 2013 program.
Rationale PhyloXML is an XML format for phylogenetic trees, designed to allow storing information about the trees themselves (such as branch lengths and multiple support values) along with data such as taxonomic and genomic annotations. Connecting these pieces of evolutionary information in a standard format is key for comparative genomics.
A Bioperl driver for phyloXML was created during the 2008 Summer of Code; this project aims to build a similar module for the popular Biopython package.
Mentors Brad Chapman
Christian Zmasek
Rationale I developed Bio.Geography, a new module for the bioinformatics programming toolkit Biopython. Bio.Geography expands upon Biopython’s traditional capabilities for accessing gene and protein sequences from online databases by allowing automated searching, downloading, and parsing of geographic location records from GBIF, the authoritative aggregator of specimen information from natural history collections worldwide. This will enable analyses of evolutionary biogeography that require the areas inhabited by the species at the tips of the phylogeny, particularly for large-scale analyses where it is necessary to process thousands of specimen occurrence records. The module will also facilitate applications such as species mapping, niche modeling, error-checking of museum records, and monitoring range changes.
Mentors Brad Chapman
Stephen Smith
David Kidd