Abstract:
Entering university, I was almost ready to give up on mathematics and fully pursue biology. Yet a growing curiosity about statistics and its real-world applications kept one foot in the door – and kept pulling. This blog narrates my journey with maths through detours in biology, a bioinformatics research project and eventually into statistical genomics. Mathematics felt abstract and disconnected until I realised its versatility. From then on, it became a slippery slope that led me deep into the realms of Statistics.
Blog:
I have always been a STEM person – but not in the tidy, focused way that sentence usually implied. My interests ranged across the full spectrum of STEM: from memorisation-heavy world of biology with its cellular machinery and genetic codes, to the clean problem-solving of mathematics where the answer either works or it doesn’t. For a long time, I managed to hold both interests together, but university has a way of forcing you to choose.
As the initial years went on, mathematics began to feel increasingly abstract and disconnected from anything tangible in the real world. There was the persistent steroretype that preceded it after all – how maths degrees leads somewhere between academia and unemployment. Biology on the other hand felt tangible, it had hospitals and drug discovery. It felt like it mattered – and I began to wonder whether mathematics was really my path.
So I decided to give it one last chance. I applied for a bioinformatics research project at the Walter and Eliza Hall Institute of Medical Research – a final line in the sand. If this didn’t convince me, I told myself I’d leave the numbers behind and go all-in on the biology. I showed up with pipelines to build and Seurat objects to process – with a vague hope that something would click.
What I didn’t anticipate was that the biology would be the overwhelming part. Surrounded by molecular biologists and immunologists who spoke fluently in pathways and protein interactions, I felt completely out of my depth. But something unexpected kept me ground – the statistics. When everything else felt like a foreign language, the statistical reasoning felt familiar in dimensionality reduction and testing for differential expression – these were frameworks that I could think in. The data had patterns and mathematics was the language for finding them.
“This is maybe not that bad” I thought. “Actually, this might be exactly the thing.”
That realisation convinced me to give mathematics another shot – and this time lean even further into the statistical aspect. I joined the AMSI Summer Research Scholarship program to work on statistical methods applied directly to genomics which meant more modelling and more algorithm development. Rather than just applying the tools, I wanted to know why these statistical methods worked the way they did.
The program had its share of humbling moments. The time the high-performance computing cluster churned for forty hours only to produce nonsensical results due to a misconfigured parameter. Moments where the model outputs look nothing like expected and the subsequent hours spent staring at diagnostic plots to identify what had gone wrong. It was not an easy journey by any means.
But two things kept me going. One was the support of my supervisor Heejung whose patience and guidance made the crashouts bearable. The other being my own stubbornness as a sore loser. Ever since I was young I never liked being defeated – not by a coding bug, not by a matrix inversion and certainly not mathematics.
The statistician John Tukey once said that the best thing about being a statistician is getting to play in everyone else’s backyard. Genetics, economics, ecology – the statistics travels where there is data. Now I can’t think of a better description of what drew me in and what now keeps me here. So I guess I’ll keep going and see whose backyard I end up in next.
Ashton Lu
The University of Melbourne
