Our Work - Data Science

Data Science

The Data Science platform facilitates the development, management, analysis, and linkage of clinical, administrative, and other data resources for patient-oriented research. This platform works closely with the Manitoba Centre for Health Policy (MCHP), health regions, Manitoba Health, Statistics Canada, and other national, provincial, and regional agencies and organizations to achieve its mandate.

Biostatistics Consulting Unit

The Biostatistics Group serves as a resource of statistical expertise for researchers and students across the University of Manitoba, as well as at other universities. Expertise is provided in design, measurement, and analysis of health data. In 2014 the Biostatistics Group transitioned into CHI as part of Manitoba's CIHR-funded SUPPORT Unit.

The Biostatistics Group serves as a resource of statistical expertise for researchers and students across the University of Manitoba. The Biostatistics Group operates on a cost recovery model, however, there are accommodations for students and faculty members who do not have funds to support their research.

The Biostatistics Group promotes two types of activities: collaboration and services.

The following activities are examples of academic collaboration. 

• Faculty/researcher/student research collaboration including research design, data collection, data analysis and interpretation.
• Research methodology development.
• Grant/funding planning, development, and proposal preparation.
• Inter-disciplinary or multi-disciplinary collaboration on research teams.

The following activities are examples of services:
• Sample size and power calculations.
• Data collections and preparation for data analysis.
• Data analysis and result interpretation.
• Database conversion, storage, archiving management and support.
• Result section preparation for manuscripts and project reports.
• Graduate student training of use of statistical software
• Survey data and survey designing.
• Statistical computing using SAS, SPSS, R and other software (e.g., HLM, M-plus).

In addition, Biostatistics Group offers non-credit courses on use of statistical computer software such as SAS and SPSS and deliver data management workshops on excel, access, database plan and design

Click HERE to learn about the Biostatistics Group and for contact information. 

Computational Biology and Bioinformatics Unit

Computational Biology and Bioinformatics Unit is a multidisciplinary team of computer science, statistics and biology.  We offer the analyses of big omics data sets for collaborative research projects and fee-for-service projects.  We provide a wide range of advice and assistance for omics studies:

  • Study design and analytical planning
  • Sample size and power calculations
  • Analysis of study data
  • Interpreting and summarizing results
  • Assistance for grant proposals

We are specialized in (1) array-based gene expression, methylation and proteomic analysis; (2) DNA-Seq, RNA-Seq, Chip-Seq and methylation sequencing analysis; (3) statistical genetic analysis;  (4) copy number variation analysis; (5) microbiome and metagenomic analysis; (6) pathway and/or gene set based analysis and enrichment map visualization.

Dr. Pingzhao Hu: Pingzhao.Hu@umanitoba.ca

Room 308 - Basic Medical Sciences Building, 745 Bannatyne Avenue

Support for New Clinician Scientists

The Biostatistics Group offers multiple training opportunities for the next generation of clinician scientists. The Royal College of Physicians and Surgeons is adding a new scholarly competency to their training standards for resident and fellows (CanMEDS 2015), which will oblige trainees to formally study research methods and conduct a scholarly or research project during their training. To support resident and fellow research, the Biostatistics Group offers the following complementary programs and services:

  • Open House sessions, where residents and fellows can meet one or more biostatistical consultants to discuss their study design and analysis questions. These sessions are co-sponsored by the College of Medicine’s Office of Postgraduate Medical Education (PGME). Registration is required, but there are no costs to participate. Please contact PGME for more information (pgme@umanitoba.ca).
  • A series of 3 academic half day workshops on statistical methods that cover the following topics: (i) study design and sample size calculations, (ii) introduction to data analysis, and (iii) statistical computing (laboratory session). Interested program directors should contact Dr. Atul Sharma of the Biostatistics Consulting Group to arrange workshop offers (Atul.Sharma@umanitoba.ca).
  • One-on-one consulting sessions. Up to five hours of no-cost consulting time is available for inquiries about study design and analysis. For more information, visit the following (http://chimb.ca/datascience#biostatisticsunit).
  • Multi-day workshops on more advanced topics, including propensity scores models for observational data, hierarchical models for longitudinal/ multi-level studies, and latent class analysis. Note that a nominal registration fee is charged for these workshops. For more information on upcoming workshops, visit the following (http://chimb.ca/events).

We strongly encourage trainees to undertake their own statistical analysis. However, for specialized analyses, residents and fellows may engage a biostatistical consultant on a fee-for-service basis. Some clinical departments may have specific funds set aside for training purposes to support the costs of a consultant. Please contact your department for further information about funding availability.

Research Data Management

REDCap (Research Electronic Data Capture) Data Management Software

REDCap (Research Electronic Data Capture) is a secure, web application designed to support data capture for research studies. It provides users with multiple features such as multi-site data entry, real-time data entry validation, audit trails and the ability to set up a calendar to schedule and track critical study events such as blood-draws, participant visits, etc. CHI has implemented REDCap, at the University of Manitoba, to provide researchers with a highly secure, centralized, audited environment to store manage and analyze research data. Please click here to learn about accessing REDCap in Manitoba. 



Current Initiatives

Data Science is engaging with a variety of partners from the University of Manitoba, Manitoba Health and the health regions to develop data resources and methods. Some examples include:

  • Provincial and health region population projections
  • Development of techniques to evaluate the quality of administrative health databases
  • Linkage of clinical research data with administrative data to study population health and health services use

Services Offered

i.      Collaborative opportunities in innovative new research areas such as:

  1. Database linkage methods
  2. Data quality assessment methods
  3. Development of instruments/tools for measurement of patient-reported outcomes
  4. Analysis techniques for complex databases
  5. Clinical risk prediction tools
  6. Diagnostic accuracy studies


ii.      Research data management

  1. Development of case report forms
  2. Data entry
  3. Data quality evaluation
  4. Linkage of databases
  5. Database storage, archiving, and management


iii.      Biostatistical consulting

  1. Sample size and power calculations
  2. Data analysis and result interpretation
  3. Result section preparation for manuscripts and project reports
  4. Training in the use of statistical software
  5. Statistical computing


iv.      Bioinformatical consulting

  1. Array and assay-based analysis
    • Differential Analysis of data from various platforms
    • Time course studies
    • Multiple experimental factors
      • Clustering and class predictions
    • Multiple hypothesis testing
    • MicroRNA differential analysis and target prediction
    • Meta-analysis across platforms and studies

  2. Next generation sequencing-based analysis
    • Sequence alignment
    • Variant calling
    • Variant filtering and annotations
    • Phenotype prediction
    • Rare variant-based association analysis
    • Differential analysis of RNA-Seq
    • Peak calling and annotation of ChiP-Seq
    • Differentially methylated region (DMR) analysis

  3. Statistical genetic analysis
    • Genetic linkage mapping analysis
      • Discrete or quantitative traits
      • Linkage analysis with covariates
    • Association analysis of quantitative and discrete traits
      • Family-based design
      • Case-control design
    • Haplotypes-based association analysis

  4. Copy number variation analysis
    • Analysis of data from different platforms (aCGH, Agilent, Affymetrix, Illumina, etc.)
    • Identify sample-specific and recurrent copy number change regioins
    • Copy number variation-based association analysis

  5. Microbiome and metagenomic analysis
    • Sequence alignment
    • Diversity and phylogenetic analyses
    • Clustering analysis
    • Biomarker discovery
    • Association analysis of host genomic variants with its microbiome

  6. Pathway analysis
    • Over-representation analysis
    • Gene set enrichment analysis
    • Enrichment map analysis
    • Module-based association analysis