Neuroscientist Unlocks Brain’s Secrets Through Gene Analysis
From advertising to academia, Michael Oldham’s journey reveals key insights into brain tumors and data reproducibility.
After a career pivot from advertising, computational neuroscientist Michael C. Oldham is transforming our understanding of the human brain’s intricate architecture using gene coexpression analysis. His innovative work offers potential breakthroughs in treating malignant gliomas and improving data reproducibility.
An Unlikely Path to Neuroscience
After initially planning a medical career, Dr. Oldham, a Duke University graduate, realized treating patients wasn’t his true calling. After a period in San Francisco’s advertising world, he returned to academia driven by a desire to understand the genetic factors differentiating human brains.
“The genetic changes that gave rise to the modern human brain were the catalyst for life as we know it,”
Dr. Oldham said. This question led him to a PhD at UCLA, setting the stage for his groundbreaking contributions.
Pioneering Gene Coexpression Analysis
Working alongside **Dr. Dan Geschwind** at UCLA and biostatistician **Dr. Steve Horvath**, **Dr. Oldham** conducted an unprecedented genome-wide analysis of transcriptional covariation within the human brain. His key insight was that recurring patterns of gene activity mirrored the transcriptional signatures of different cell types.
Published in Nature Neuroscience in 2008, his work demonstrated how gene coexpression analysis could pinpoint optimal markers for cell types and states. This principle remains central to his lab’s research at UCSF.
Weighted Gene Coexpression Network Analysis (WGCNA), the method he developed, has become crucial in genomics research. The approach identifies coordinated gene activity patterns within biological systems, making it invaluable for studying complex tissues like the brain. Current analysis utilizes AI to analyze complex patterns, in line with global trends (Nature, 2023).
From Evolution to Tumors
Dr. Oldham’s early research focused on gene activity patterns in human and other species’ brains. He identified gene expression changes in human radial glia, interneurons and astrocytes. More recently, his focus has shifted to malignant gliomas, applying his computational expertise to these challenging brain tumors as part of UCSF’s Neurological Surgery department.
His team has analyzed gene activity patterns from over 17,000 human brain samples, leading to OMICON (theomicon.ucsf.edu), a platform providing access to gene activity patterns in complex datasets. The resource offers researchers insights into brain function and dysfunction.
Targeting Gliomas
By comparing gene activity patterns between healthy brains and malignant gliomas, Dr. Oldham‘s team is identifying molecular changes in glioma microenvironment cell types, like vascular cells and neurons. These signatures may lead to novel biomarkers and targeted treatments, using cell-surface markers of glioma vasculature as a potential “zip code” for bloodstream targeting.
Addressing the Reproducibility Crisis
Dr. Oldham has also turned his attention to science’s reproducibility crisis. As Vice Chair of UCSF’s Academic Senate Committee on Library and Scholarly Communication, he launched a task force on research data and metadata standardization.
“If most of the findings we toil to produce cannot feasibly be reproduced, what is the point?
he asks, underlining the need for more open and reproducible science.
“Although there are many factors that affect the reproducibility of published research findings, there is no reason in principle why data analysis should not be completely reproducible,”
Dr. Oldham stated, advocating for standardized data packaging to accelerate data discovery and analysis.
Personal Reflections
The interview also reveals the human side of Dr. Oldham’s scientific journey. His choice to spend two extra years in graduate school, considered unconventional at the time, led to a significant paper that secured his UCSF Sandler Faculty Fellowship.
Outside of neuroscience, Dr. Oldham enjoys hiking in Marin County and maintains friendships from his advertising days, living by the motto: “ABC (always be celebrating!).”
Looking ahead, Dr. Oldham emphasizes integrating multiscale and multimodal data and using robotic automation to generate reproducible datasets. He also advocates for shifting from descriptive to predictive analysis using statistical models.
“There is a big difference between describing what you think a dataset means versus predicting what you will see in the next dataset,”
he concludes.