Note: This post is going to be a bit more personal, so if you don’t care to read my ramblings on what I’m currently doing and what I’m planning for the next few months, just give this one a skip and I promise to have more data-related goodness for the next post. *

About a month ago, I found out that I was being let go from my last position as Field Application Specialist for a lab equipment manufacturer, which prompted me to finally go ahead with my long-planned reskilling into computational biology.

As I mentioned in the about page here, I decided some years ago that I didn’t want to be doing lab work at the bench full-time for my whole career and started to embark on projects that had increasing amounts of computational analyses. But, in my whole education as a Molecular Biologist, I never had any formal training in programming and I had to pick up the skills I do have now on my own, as needed for each project. This means I have done a fair bit of scripting in Python, MATLAB, Perl, and Bash but I never had to do any of them for long enough to become fully comfortable with any of these languages.

At my most recent research position, I also developed a strong interest in machine learning, more specifically on how it’s increasingly being used in academic and industrial biological research. So I joined this interest with the desire to spend more time programming and decided to take courses that would help fill some of the gaps in my training and also give me additional tools to pursue fun personal projects. I decided to focus on making python my programming language of choice, since it is increasingly becoming the de facto standard for machine learning applications.

The first course that I picked was the very well-recommended “Machine Learning” by Andrew Ng on Coursera, which I actually started a few months ago and already completed. While I was not a big fan of the programming exercises in Octave/MATLAB, his teaching style based on digging deep into the maths behind the various techniques shown in the course was a great way to get a good understanding of how lots of fairly simple calculations can lead to the awesome results being produced with machine learning techniques nowadays!

The next course, which I already mentioned in the previous post, is Jeremy Howard’s Practical Deep Learning for Coders, using the fastai library that sits on top of PyTorch, making it very quick to produce working deep learning models while also providing access to lower-level controls for fine-tuning and improvements. I’m still in the first half of this course, but I’ve very much enjoyed the lectures and exercises that I’ve completed so far, so I think this course and the recently-released book will end up as very useful exercises in this journey.

For the other course I will soon tackle, by total chance I recently happened upon a Twitter post mentioning an online reading group for a book on Bayesian statistics, so I searched around a bit and decided to join the group! The book is Statistical Rethinking by Richard McElreath and there are accompanying videos of the course taught at the Max Planck Institute in Leipzig. I’ve been keen to learn a bit about Bayesian stats for some time, especially with all the recent opinion papers claiming that the traditional frequentist stats are not ideal for research. The package used by this book is R-based, which is not ideal given my plan to focus on Python, but R is also very widely used (probably even more than Python for some visualisation and stats applications), so it is useful to know!

The very fortuitous way that I found the reading group for McElreath’s book and how that changed my plans reminded me of this song from The Lord of the Rings that I think about sometimes: * *

The Road goes ever on and on
Down from the door where it began.
Now far ahead the Road has gone,
And I must follow, if I can,
Pursuing it with eager feet,
Until it joins some larger way
Where many paths and errands meet.
And whither then? I cannot say.

– J.R.R. Tolkien (in the voice of Bilbo Baggins)

While I have had this plan to reskill into computation for quite a few years, I’ve been letting “The Road” sort of dictate how I get to that end goal. That road has been a very winding path until now and I have certainly enjoyed the majority of that time, but this most recent fork that I’ve encountered seemed to present a much more direct route, which is the option that I chose this time.

Footnotes

* I’ve just realised that with this post, I’m now on the fifth instance in a streak of roughly one post every 5 days. In the interest of keeping this nicely arbitrary trend going, I’m going to make this the update frequency for this blog. It will give me a quick metric to track and also act as a good way to hold myself accountable!

* * One of the very few songs in the books that I really bothered reading, despite being enough of a Tolkien fan during school that I read the entire trilogy 5 times!