Education and Outreach
The Broad ICBP/CCSB center has a diverse array of Education and Outreach activities that share a common overarching goal: to train the next generation of cancer systems biologists. We aim to accomplish this goal via four categories of activities:
- Developing and disseminating advanced yet accessible computational tools
- Hands-on training at the interface of cancer biology and computational science
- Producing and disseminating videos for the research community and general public
- Sponsor high-school students and post-docs from underrepresented backgrounds
These activities embody multiple levels of Education and Outreach. First, our Broad ICBP/CCSB Center relies in large measure on havinga fully integrated cancer biology and computational biology effort and the interdisciplinary training we provide is critical to our success. Additionally, the success of the NCI ICBP/CCSB network as a whole relies on Outreach activities that connect Centers to each other, Outreach activities that connect Centers to the greater scientific community, and Outreach activities that connect Centers to the general public (click here for more details). Our program accomplishes each of these levels of Outreach, complementary to larger Outreach initiatives at the Broad Institute.
1. Developing and disseminating advanced yet accessible computational tools
We have developed and continue to develop a set of a set of caBIG-compatible computational tools that are powerful while being accessible to biologists. These tools are widely adopted by the scientific community and now have over 50,000 users from around the world.
Not all researchers can attend our workshops and therefore for all of our software tools we provide an informational web site. At these sites users can download the software, and find information about the tools, including user guides, analytic protocols, and online tutorials.
All of our software is available here.
2. Hands-on training at the interface of cancer biology and computational science
Computational workshops for biologists
Over the past 5 years, we have trained nearly 500 biologists in the underlying mathematical and statistical principles underlying the analysis of gene expression data. These workshops are held 4–6 times a year and are of varying length — from 1–3 days. In addition to explaining the theory behind the methods, we encourage participants to bring their own data for hands on sessions where they use GenePattern or other relevant tools for analysis. The syllabus for the expression analysis workshop includes units on: (1) Data Acquistion, (2) Data preprocessing, normalization, and filtering, (3) Differential Analysis — finding markers or differentially expressed genes, (4) Estimating significance, permutation testing, and multiple hypothesis testing, (5) Discovery and Clustering (hierarchical, NMF, k-means, probabilistic and mixture modeling, consensus clustering), (6) Prediction or classification (weighted voting, naïve Bayes, shrunken centroids, SVM, large Bayes, K-NN), (7) Cross- and external-validation, (8) Cross-platform analysis, (9) GSEA, (10) Computational Genomics Tools — Bioconductor, GenePattern, Matlab, DChip, and others. In addition, we have run tutorials for SNP analysis including such advanced algorithms as GISTIC.
We will continue to run these tutorials and workshops for the research community and develop and add units that are particularly relevant to cancer systems biology and the proposed new directions of our Center. New units will include (but are not limited to: 1) GENE-E tutorial — theory and practice of analyzing and viewing RNAi screening results; 2) IGV tutorial; 3) RNAi scoring and analysis — theory and practise. 4) Connectivity Map workshop — how to interpret the results of a C-Map query; 5) FLAME — the theory and practice of modeling flow cytometric data; and 6) Predicting essentiality from molecular features — theory and practice.
In addition to these face-to-face workshops, we plan to (1) produce videos of the most popular ones that can be streamed from the Web and (2) post all curricula and course materials online so that the workshops can be replicated at other sites and institutions. We will also create very short — 2–3 minute flash demos of all the software that we develop and disseminate related to our Center.
Laboratory workshops for computational scientists.
Mathematicians, physicists, computer scientists, and software engineers can approach computational biology problems with great intellectual rigor and precision and can bring to bear extraordinary computational power for integrative cancer biology research. Often, however, researchers from quantitative backgrounds do not fully appreciate the experimental difficulties in generating biological data, and especially the issues of reproducibility and noise that are central to such data.
The purpose of the laboratory workshops to be conducted by the Center will be to provide computational biologists with a hands-on understanding of experimental systems biology data sets. This training will take place as an intensive 2-week long experimental workshop led by experienced Center scientists. The curricula for these laboratory workshops will be packaged as distributable coursework via the Internet so that other institutions can execute these workshops as well. We will develop the details of the workshop over the first year of our grant, and pilot it with computational scientists at the Broad. By completing this workshop, students will learn to: a) Use previously generated genomic data to formulate hypotheses about essential genes in cancer cell lines; b) Design and execute RNAi experiments to test these hypotheses; and) Analyze data and present findings to colleagues in a group meeting setting.
A schematic of this workshop is shown below:
A full day by day agenda can be found here.
3. Produce and disseminate videos for research community and general public
Part of our outreach philosophy is to connect in a meaningful way with as much of the research community as possible. While our workshops promise to be effective, they will reach only a small number of individuals. To broaden our outreach, we will produce, and disseminate via the Web, videos by experts in the field describing the opportunities and challenges in cancer systems biology. This represents a new direction for the Center, but one worth exploring. In addition, in recognition of the need to educate the public about the emerging field of cancer systems biology, we will produce multi-media presentations utilizing new Microsoft Surface technology, as described below.
Drawing on the resources of our own Center, the Boston community, and the other CCSBs across the country, we propose to develop short videos, at most 15 minutes in length,that can be streamed over the web. These videos will be mini-lectures or chats featuring scientists who are actively working in cancer systems biology and well recognized for their work and accomplishments. The topics will be up-to-date, and will aim to capture both the excitement of cancer systems biology, new experimental or computational approaches, and key challenges. We will focus on our own Center faculty and senior staff for initial videos, and then will expand to include expert faculty outside the Center.
Multi-media presentations for the lay public.
Our public outreach program will take advantage of the Broad’s DNAtrium located in the institute lobby. In another experimental effort, we will take advantage of the Microsoft Surface Units installed at the Broad. These table-like surface platforms respond to natural hand gestures and real-world artifacts (e.g., laboratory devices) containing embedded radio frequency ID tags. It is simple, fun, intuitive, multi-user experience. The Surface exhibits will be designed to appeal to the general public, allowing people to discover the importance of cancer systems biology and its potential for long-term impact on human health. We will pilot two projects for the Surfaces in the form of virtual laboratory experiments to provoke the imagination of visitors and inspire a next generation of cancer scientists (and their parents). Specifically, we will produce presentations on 1) The Connectivity Map, and 2) Interrogating biological networks using RNA interference. If these are successful, we will explore deploying other ‘experiments’ relevant to the work of our Center. Importantly, we will make the code for these presentations available to other centers with access to Microsoft Surface technology.
4. Sponsor high-school students and post-docs from underrepresented backgrounds
The Broad Institute has diversity initiatives aimed at increasing the representation of underrepresented minority groups in the field of genomics and cancer research. These programs include the the Summer Research Program in Genomics (SRPG) for undergraduate students and the Minority Introduction to Engineering, Entrepreneurship, and Science (MITES) at MIT. Both are rigorous summer programs introducing high school students and undergraduates to engineering, science and entrepreneurship. For minority college students interested in obtaining research experience, and minority students interested in pursuing graduate degrees in bioinformatics, the Broad has developed a formal postdoctoral training program in genomics aimed at underrepresented minority students, as well as a collaboration with the Harvard–MIT Division of Health Sciences & Technology’s Bioinformatics and Integrative Genomics Program (BIG). As with the prior funding cycle, we will aim to have 1–2 such students join the Center each year. For details on each of these programs, please visit the Broad Institute Diversity Initiative in Scientific Research.
Additionally, we have hosted and are continuing to host summer interns each year that are selected to participate in the NCI ICBP Summer Research Program. We have included them with other Broad summer students to provide a peer cohort. They attend all of the same introductory seminars and social functions and are housed in the same location. In addition, each Center investigator mentors undergraduate and/or graduate students from either Harvard or MIT, and these students generally sit together in the laboratory — whether they are experimentally or computationally oriented, so as to maximize interactions at the interface of cancer biology and computational biology.