Skip to main content

BF591 - R for Biological Sciences

Assignment 1

Problem Statement

This assignment will focus on basic functions of R with an emphasis on tidyverse implementations. tidyverse is a collection of packages, pioneered by Hadley Wickham and RStudio, that looks to standardize procedures, functionality, and syntax in R.

To gain familiarity with R, we will be working with a microarray dataset that contains gene probe expression data for various samples collected from cancer patients. In bioinformatics, it is common to have multiple datasets for your different modes of data (i.e. microarray expression data is kept separate from clinical data detailing the samples). You will get an opportunity to work with both of these datasets, and be required to cross reference between the two.

Learning Objectives and Skill List

  • Install various packages needed for analysis
  • Load Data
  • Gain familiarity with common tidyverse operations such as groupby(), mutate, and summarize.
  • Create a small plot to display results
  • Utilize R Markdown to create an attractive format for sharing data.

Instructions

Our main focus for this assignment are installing packages, manipulating data, and summarizing important statistics across both samples and features (genes). An empty project repo can be found here:

https://github.com/bu-bioinfo/bf591-assignment-1

The project is laid out as such:

main.R
report.Rmd
reference_report.html

Note that in Assignment 1 there is no test_main.R and does not use the unit testing framework as described in the Assignment Overview. For this assignment you should just compare your knitted report to reference_report.html. All subsequent assignments have test_main.R.

Each step of the assignment is explained in the R markdown file, report.Rmd. There you will find a list of tasks to explicity implement functions in your empty main.R script. The main.R script contains skeletons of each function you’ll need to implement, explaining what each function should do, the parameters it expects to receive, and what type of output is expected to be returned. A reference report, reference_report.html is also provided. Assuming you successfully implement all the functions in main.R, your generated report should look identical to the information displayed in reference_report.html. In this way, you can use reference_report.html as a guide to determine if you are correctly implementing your functions.

Here is the suggested workflow for developing and checking your code in this assignment:

  1. main.R contains function definitions, including signature descriptions, for a number of functions, but the bodies of those functions are currently blank
  2. report.Rmd has code chunks that call functions defined in main.R - you do not need to write anything in the Rmd file (but you may)
  3. Your task is to read the function descriptions and the text in the Rmarkdown document and fill in the function bodies to produce the desired behavior in main.R
  4. You can test your work by executing individual code chunks in report.Rmd and comparing your output to the example compiled report in the repo
  5. In the workflow, you will go back and forth between developing code in main.R and running code chunks in report.Rmd
  6. When you have developed function bodies for all the functions and executed all the code chunks in the report successfully, you should be able to knit the entire report

Hints

  • When developing the period_to_underscore() function, you might find the stringr::str_replace_all() function helpful. The pattern argument to these functions is interpreted as a regular expression or “regex” for short. A regular expression is a sequence of characters (i.e. a string) written in a language that describes patterns in text, similar to “Find and Replace” operations in word processing software, but is more powerful and flexible in the kinds of patterns it can detect. Some characters have special meaning in regular expressions, one of which is the . character. In order to identify the literal period character like we are trying to do, we must instruct the regular expression to do so by either escaping the character with \\. or place it in a range with [.]. Either of these two methods will work to replace a literal . with _. See the section on Regular expressions for more information.
  • If you are getting type conversion errors when loading in your expression CSV file, check to make sure you aren’t supplying your own column names. The CSV file has column headers already, and supplying your own will cause the first line to be read in as data. Since the first row are character values in this case, all of the other values in the columns will be coerced to characters as well, instead of reading them in as numbers.