Skip to main content
BrainWaves: The Neuroscience Graduate Program Newsletter

MiNDS Manuscripts: Jee Su Suh

Author: Claire Gage

This month in the MiNDS Manuscript series, Jee Su Suh discusses her paper recently published in the journal PsychoneuroendocrinologyClick here to read the article!

Hypothalamus volume and DNA methylation of stress axis genes in major depressive disorder: A CAN-BIND study report

What was the primary focus of your paper?

We wanted to investigate the association between epigenetic markers in stress axis genes and hypothalamus volume in living humans diagnosed with major depressive disorder. Using an age-matched healthy control sample, we wanted to see whether:

  1. There was an observable relationship between the brain and blood-derived DNA methylation
  2. This relationship was altered in depression

Can you briefly explain the findings of your paper?

I used an elastic net model (incorporating simultaneous feature selection and linear modelling) to investigate whether % methylation of various stress axis genes could predict hypothalamus volume. It turns out that methylation was more predictive of hypothalamus volume in depression than in controls. This was interesting given that group comparisons of either volume or methylation on their own were not statistically significant. It was only the relationship between these variables that differed between groups.

How did you develop this research question?

I had previously worked on whether we could use MRI data to predict clinical outcomes in depression; specifically, the outcome of antidepressant treatment. It turns out, most simple associations between brain measures and clinical outcomes are not statistically significant. For this analysis, rather than focusing on which single variables differed between groups, I focused more on whether the associations between various data sources were different. This was the main inspiration behind looking at brain-behaviour relationships within each group and doing more of a qualitative comparison between them, rather than jumping ahead to statistical group comparisons of single variables.

What were some issues you faced when writing this paper? How did you overcome them?

The first submission of this paper was rejected with some brutal but incisive peer reviews. They pointed out some fundamental errors in reasoning and analysis, which stung at the time but for which I am now grateful. I took some time off from the paper -- conveniently, this was around the winter holidays, so I went a good 3 weeks not thinking about it. Once I went back, I started all my analyses from scratch and felt almost as though I coasted to the final results that are published today. What I’ve found helps a lot is to start from first principles whenever I am in doubt, and a mental blank slate is tremendously helpful. It was also great to have my supervisor’s support to take the time I needed to clear my head and the encouragement from lab-mates and friends throughout the process.

What does your paper hope to accomplish/ how does it add value to the field?

I think it’s interesting to see the alteration of biological relationships within depression. The results highlight that even if we can’t necessarily distinguish more obvious differences between groups (e.g., participants with depression exhibit smaller/larger volumes), there are detectable changes at a higher-level biological level (i.e., methylation and brain volume are more tightly associated in depression). It also reminded me of how many biological relationships we have yet to investigate within depression, which I think would go further towards clarifying the sources of clinical heterogeneity between patients.

Do you have any advice for anyone who is writing their first research paper?

The situation I described above has emerged as a pattern throughout my graduate career. I used to feel terrible whenever I received any negative feedback in the form of peer reviews. Now, after a few papers, I have come to expect that my initial submission will likely have errors regardless of how long I spend working on it, and that it is only the first step among many iterative improvements. I have come to be grateful for such “negative feedback” because they signal concrete changes I can make for the better. Going into my 5th year of graduate studies, I’ve learned to rest easy in the knowledge that as long as you do your work in good faith, following the highest standards of analysis to the best of your ability, things will turn out fine regardless of the results.