How to start analyzing your own 'omics data (without coding)

Effective, precise and personalized diagnostics and treatment stemming from an improved understanding of diseases has been in part driven by increased and improved data collection, especially high-throughput data. Genomic, transcriptomic, proteomic, metabolomic, and other ‘omic datasets enable biological research on a scale not possible even in the recent past.

While today’s massive and exponentially growing body of data provides an unprecedented level of detail about molecular mechanisms associated with disease, the generated datasets are huge, heterogeneous, full of artifacts, and very complex. Realization of the potential of these resources - whether in basic science research, translational research, biotech, or clinical practice - requires practical education in the technological skills needed to harness them. Such education must go beyond theory to provide a practical understanding of the tools, approaches, logic, expected outcomes, opportunities, and applications needed to handle, analyze, and interpret such datasets.

We recently started asking our contacts (https://goo.gl/forms/8VQhTRqrafZh6eFo1) how important was it for them personally to be able to analyze and interpret omics data. While the survey is still ongoing, a wide range of people already responded - professors, Ph.D. students as well as healthcare and biotech professionals. But most tools require one to be comfortable coding and a good understanding of the logic of big data analysis.

While computer coding literacy has increased substantially among students, researchers, and clinicians, most are not ready to use code line interface and/or are not inclined to invest the significant time needed to learn this skill. While solutions using a graphic user interface have started to appear, most of these require a good understanding of input/output and configuration of algorithms as well as the logic of constructing pipelines of algorithms. In addition, working with big data requires a foundation in statistics and machine learning as well as substantial computational resources, which are expensive and require hardware expertise to assemble. All these factors hinder the usefulness of available public domain datasets and tools, limit the active adopters to those who have access to resources, and delay the adoption of big data for use in biomedical practice.

This is especially apparent farther away from established clusters of advanced universities and high-tech centers, located disproportionately in the US Northeast and California. Such institutions are supported by large grants and higher incomes, and therefore already have the necessary resources and talent. In contrast, more isolated communities, are characterized by a lack of experience and availability of such resources and skills. This is also true outside of the US.

Seeing this need, we decided to develop a set of practical, modular courses in ‘omics data analysis based on public domain projects using our user-friendly, web-based bioinformatics analysis platform. The goal is to skip the complexities of coding and theory, jumping right into practice. Our goal is to start a series of hands-on workshops and make all the data accessible online. Updates coming soon, reach out to me if you want to participate! [email protected]