Learning in this site
Just starting? Trying to figure out if this site will be relevant to you? A good place is to begin with the glossary of terms, and then choose “Arcus Orientation” as your “educational pathway” below to get a selection of articles that probably makes the most sense for what you need.
Also, check out the various specific educational pathways here, for the researcher who wants to assemble a meaningful series of things to read and do. Please mix and match what makes sense for you!
Curious about Arcus?
Your first stop in understanding Arcus is to grasp why we exist and how we want to help your science advance!
You may also want to understand a bit more about Arcus services:
Tabular Phenotypic Data Researcher (the Excel Guru)
Do you work with data that can be displayed appropriately in a table (like a spreadsheet in Excel)? Are your data files on the smaller size (< 100k rows, <500M file size)? You're a perfect candidate to learn more about using reproducible tools to prepare and analyze your data. If you do most of your analysis in Excel or in a program like SPSS by using point-and-click (not syntax or code), you'll want to get up to speed on a few topics quickly:
You're used to doing advanced work on your own data, but you're suddenly faced with trying to integrate, clean, and prepare data from several sources, possibly including large clinical datasets. You need to be able to "munge" (reshape prepare for computation) a bunch of data quickly and reliably. Try:
It's been awhile since your last statistics class, and you need to brush up on basics like null hypothesis testing, correlations, linear models, effect sizes, and picking the right statistical test for your experiment. Check out these articles!
Poster Maker / Visual Thinker
You're statistically solid and don't need much help with the analysis portion of your research, but you would love to be able to generate better publication-quality graphics (or even just exploratory visualizations for yourself!). You would do well to check out:
You work with data that is interconnected: family members, linked by common heredity, community members connected by exposures to contaminants at particular locations, scientific authors who are linked by collaborations, dimensional aspects of phenotypes that are connected to genes or exposures, etc. You want to use elements of graph theory to store, visualize, and analyze your data. Start here!
You work with data that is big -- large files (measured in gigs). You need high performance computing and advanced methods. You'll want to know about:
Predictive Model Builder
You already use a statistical programming approach to prepare, analyze, and visualize your data. Now you want to learn how to use machine learning to come up with predictive models based on your data. Flogging Stack Overflow is not how you want to spend your day. You may be interested in:
You've assembled data that consists, at least in part, of natural language. You want to learn how to analyze texts by sentiment, part of speech, keywords, etc.