Bioinformatics & Biostatistics Unit

Goals and Objectives: To provide the statistical, analytical and computational training and support needed for all MSU researchers for data analysis and research publication.

Plans: The role of sound experimental design and data collection, analysis, and presentation in biomedical research, and especially in data intensive fields such as social and behavioral sciences, genomics, proteomics, and computational biology, cannot be overemphasized. Given the complexity of modern biomedical research data, a biostatistical and computational support unit is necessary to assist at the different stages of the proposed pilot studies and other biomedical research projects at MSU. This unit will provide technical consultation to investigators in all phases of research including design, data collection, filtering noise and extracting informative information, data exploration using different statistical approaches, and developing tools, methods and packages required to automate and facilitate data analysis and interpretation. This unit will also provide training workshops, short and regular courses in biostatistics, bio-programming, bioinformatics, and computational biology with a focus both on genomics and proteomics, for investigators, researchers, students, and project collaborators.

Increasingly, raw biomedical data with the attendant meta data are becoming available to the research
community and it is envisaged that the unit will work closely with other investigators to develop strategies for
data storage and sharing and to develop data mining pipelines for both research and student training. The unit
will assist investigators in proposal development by providing sound statistical and computational rationales for
study design.

The BBSU will be equipped with three workstations for training and research. The following software will be purchased with site licenses for installation in any workstation on campus: 1. atlas.ti; 2. NVivo; 3. STATA; and 4. SPSS extended version. The SPSS software package is currently available at MSU and is widely used in data analysis because it incorporates many basic and advanced statistical routines. It is very well suited for quantitative data analysis as derived from laboratory and behavioral studies. Qualitative data, for example from surveys, will be analyzed with NVivo, while other sources of data, such as textual data, will be analyzed with atlas.ti. We will also offer training workshops on how to use Magic and Galaxy webservers to conduct tasks from DNA and RNA sequence alignment to pathway analysis and protein function prediction. We will also use national computing resources such as XSEDE (the National Science Foundation funded program) for applications requiring a supercomputer. We will mainly focus on open source, developing platforms such as python and R programming languages to build our packages and toolbox to facilitate accessibility and usage.

Contacts:

  • Dr. Mian Hossain (mian.hossain@morgan.edu), an MSU Biostatistician, will work as the Co-Lead of the BBSU, to provide training, biostatistical, and analytical support needed for the program. Dr. Hossain will provide biostatistics training for users through workshops, short courses and training modules every semester and as needed.
  • Dr. Abdollah Dehzangi (abdollah.dehzangi@morgan.edu), an MSU Bioinformatician and Computational Biologist, will assist Dr. Hossain to provide training, back up, and expertise in the fields of bioinformatics and computational biology. As the program coordinator of the MS in Bioinformatics program and an active member of the MSU Computational Biology and Bioinformatics research core, he will provide the BBSU with expertise required to facilitate data analysis, interpretation, and training. In addition to providing training in bio-programming, computational biology, and bioinformatics tools and databases, he will also actively contribute to the development of tools and packages to be used for data analysis, interpretation, and automation of tasks using cutting-edge knowledge in machine learning with an emphasis on deep-learning architecture (see also the training section below).