MicHanThi is a software tool for the prediction of gene functions. It was designed to facilitate the genome annotation process by providing a rapid, high quality prediction of gene functions. It clearly out-performed the human annotator in terms of accuracy and reproducibility of annotations during the initial evaluation phase. Today, it is used in most genome annotation projects supervised by the Microbial Genomics Group at the Max Planck Institute in Bremen, Germany. It is also applied by Ribocon in their industry grade annotation projects.
Sequencing has become an integral part of every-day molecular biology. Even small and mid-sized groups start to apply these techniques on a regular basis. At the same time, modern sequencing technologies generate enormous amounts of (meta)genomic data in a very short time frame. Tools which can handle these data are indispensable. Most tools available today focus on initial tasks like ORF calling and similarity searches. Tools that use these results to predict a function for the potential ORFs are normally included in larger systems that automatise each step from the initial ORF calling to the prediction of gene functions and metabolic reconstruction. Also, most of these systems are web based and the control over the annotation process is limited. Most system do not offer control over what analysis tools to run and what parameters should be used for each tool. Additionally, these systems mostly cannot be installed locally. The limited influence on the annotation process as well as the missing option to install these system locally limits their use for commercial projects.
Like most UNIX tools, MicHanThi aims at solving a single problem efficiently. If a more complicated task, like the annotation of a genome, is to be achieved then several of these tools are combined, each processing the output of the previous tool(s). In particular, MicHanThi uses the results provided by tools such as glimmer, critica, BLAST, and InterProScan as input data to predict the function of a gene. Customised annotation systems (pipelines) can be easily implemented around MicHanThi when tools for ORF calling and similarity perceed MicHanThi. These systems offer complete control over every aspect of the annotation process (from the initial ORF calling to the metabolic reconstruction and comparative genomics).
Initially, a connection to the GenDB annotation system (version 2.0 and 2.2) was implemented. Support for other annotations system can easily be achieved if the `DataSource' interface is implemented. It is also possible to implement a DataSource that runs the BLAST and InterProScan tools for each ORF. MicHanThi also abstract the similarity search tools used during the annotation process. As the with data source an interface (`Tool') needs to implemented in order to support additional tools.
The Software is completely written in Java. At least version 6 (1.6) of the JAVA standard edition (SE) is required to run the software locally. Additionally the following libraries are required:
The software is available free of charge and can be dowloaded using the link provided here. If you are interested in using the software in your project and you need help please contact either Frank Oliver Glöckner or Christian Quast at the MPI - Bremen.Download: MicHanThi (1.5.0)
The original master thesis (MicHanThi - Design and Implementation of a System for the Prediction of Gene Functions in Genome Annotation Projects) written by Christian Quast can be found here
An updated publication is currently in preparation.