Essentials

1. Carefully read the Colautti Lab Manifesto

  • Essential reading for all members of the Colautti Lab

2. Learn how to research

3. Develop a project

  • Define your question and research the relevant theory
  • Identify your main hypotheses, generate quantitative predictions, and plan out your experiment(s).
  • Run your ideas by everyone who will listen, especially your mentor. Give it your best effort and then be prepared to change it.

4. Read the BES Guides to Better Science

5. Collect Data

A good rule of thumb: imagine somebody reading over your shoulder who is not familiar with your work. Would they be able to interpret your filenames, meta-data, README files, data field names, documentation, etc? A bonus is that you will be able to understand what you did when you revisit your data months or years later.

BEFORE you start

Organize and label samples carefully

Create a Data Management Plan (DMP) * Know this now – you will have to publish your data publicly, so take a bit of extra time now to make sure it is understandable. * Using short-hand names (e.g. Wt instead of total dry biomass) will save a lot of coding headaches. * Create a metadata.txt file describing the basics of your data. A short paragraph of where it comes from, and then a line describing each file and variable in your dataset.

WHILE you collect (or simulate) data

Collect and Manage Data CAREFULLY

  • Build in redundancy. Never have your data in only 1 place.
  • Collect data by hand, using paper and a pencil. For larger experiments, print out data sheets with defined cells.
  • Enter handwritten data onto a computer and back it up online. Do it WHILE you record data, not after.
  • Manage data in one or more spreadsheets, or setup a database with PyTrackDat
  • Don’t worry about trying to put everything in one spreadsheet. As long as you have ID codes (e.g. from baRcodeR) to link different experiments/measurements then it is easy to connect these later in R.

After Finishing

Publish your data

  • For data collected at QUBS, use the QUBS Dataverse Repository, open to all
  • For other datasets, use a repository like datadryad.org, FigShare.com
  • For larger collaborative projects, create a Dataverse Repository
  • Use specialized datasets like NCBI SRA database or one on the list in this paper, or search Google for a more appropriate home)
  • Consider whether it would be useful to combine repositories e.g. upload SRA files to NCBI and include Accession numbers in metadata posted in another repository containing code and called SNPs.

7. Start Coding

READ Google’s R Style Guide

Set up R Studio (Free Desktop Version) with a GitHub account on your computer.

Develop Your Coding Superpowers

Conventional Superpower Coding Superpower
Enhanced abilities Fundamentals of R and Python
Mind Control Bash and Linux Command Line Computing
Robot Army Flow Control in R and Python
Clone backup Git and GitHub
Illusionist Custom Visualizations with qplot and ggplot
Telekenesis Custom Functions in R and Python
Transmutation Regular Expresssions in R and Python
Mega-mind Resampling & Simulations in R Pt.1 Pt.2 Pt.3
Master Builder Custom R Packages

Complete Your Skillset

Here is a current, comprehensive list of guided self-tutorials, organized by theme.

NOTE: Many of these links include multiple tutorials, which are linked at the top of the webpage.

Other source material

Graham Coop’s PopGen notes

Other resources

Public Website

  • bit.ly/colautti – Typical departmental website with basic research information

Publication Record

Public Engagement

Private Lab Website