<aside> <img src="/icons/list_gray.svg" alt="/icons/list_gray.svg" width="40px" /> Home | Research Aims | My journey

</aside>

Home

<aside> <img src="/icons/mail_gray.svg" alt="/icons/mail_gray.svg" width="40px" /> Email

</aside>

<aside> <img src="/icons/follow_gray.svg" alt="/icons/follow_gray.svg" width="40px" /> @KlieAdam

</aside>

<aside> <img src="https://s3-us-west-2.amazonaws.com/secure.notion-static.com/911fee6a-b4c8-408c-bab9-97a18ece4d72/github.png" alt="https://s3-us-west-2.amazonaws.com/secure.notion-static.com/911fee6a-b4c8-408c-bab9-97a18ece4d72/github.png" width="40px" /> GitHub

</aside>

<aside> <img src="https://s3-us-west-2.amazonaws.com/secure.notion-static.com/7e8fa7d2-6aab-431e-994a-e9fcf34b7b12/scholar.png" alt="https://s3-us-west-2.amazonaws.com/secure.notion-static.com/7e8fa7d2-6aab-431e-994a-e9fcf34b7b12/scholar.png" width="40px" /> Scholar

</aside>

Ph.D. Candidate

Bioinformatics & Systems Biology

I’m a Ph.D. candidate in the Bioinformatics and Systems Biology (BISB) program at the University of California, San Diego.

My research focuses on understanding how gene expression is regulated (see below for my specific research aims).

🎓 Education

• B.S. in Bioengineering: Bioinformatics, 2017

Curriculum Vitae

CV

Research Aims

The completion of the Human Genome Project in 2003 provided a nearly complete map of human DNA, setting the stage for the field of genomics to blossom. This milestone enabled scientists to explore differences in our DNA sequence (genetic variation) by comparing genomes from many different individuals.

Today, we have genome sequences for over half a million people, identifying hundreds of thousands of genetic variants linked to common diseases like heart disease, Alzheimer’s, and diabetes. Yet, understanding the biological mechanisms behind these links remains a challenge. A significant portion of these genetic variants occur in the regions of the genome that do not produce proteins (called the non-coding genome) but instead play a pivotal role in regulating protein production in cells. Despite its importance, the non-coding genome is far less understood than the regions that code for proteins.

In my thesis, I use machine learning to analyze large-scale genomics datasets, aiming to predict how genetic variation in the non-coding genome impacts biological function.

Building a software ecosystem for machine learning in genomics

When I started my PhD in 2019, machine learning (ML) was already making a substantial impact in genomics. However, successfully applying ML was and still is consistently more challenging than is reported in the latest papers. Why?

• Steep learning curves are everywhere. Genomics involves specialized assays, unique file formats, and buggy software. Rapid exploration of data and ideas are essential for effective analysis, but even simple tasks can take several steps (and tools) to complete.
• Pitfalls are prevalent [1]. ML pitfalls, though becoming increasingly well-documented, can be difficult to detect. If not recognized, they lead to datasets that get undervalued or overestimated modeling capability.
• Model interpretation is difficult. In many domains, we can get away with a model that just makes good predictions. This is not enough in biomedical research. We need to understand how a model is making its predictions. There are now several popular methods for interpreting ML models [2], but their complexity often keep the model opaque.
• Reproducing results is challenging. Reproducibility is fundamental to scientific research, but can be difficult in ML. In genomics, this is amplified by difficult to understand method’s sections, mountains of custom code, and poor data sharing practices. These issues often make replication of results impossible.
• Best practices are evolving. Keeping up with the best practices in ML for genomics is like trying to follow the plot of The Mandalorian. It is crucial to consolidate, reconcile, and document existing methodologies for broader adoption and discourse. This process is consistently undervalued.
• No unified framework currently exists. Machine learning frameworks like scikit-learn, PyTorch, and Keras are wonderful and widely used, but they are too general for many tasks in ML for genomics. Genomics-focused tools are plentiful, but can be difficult to use and extend, either on their own or in concert with each other.

I set out…

Tools

Docs | Publication | Preprint | GitHub

Talks

https://www.youtube.com/watch?v=47wbTR9yUpg

Learning the grammar of enhancer function

An exciting direction for ML in genomics is in studying genetic switches called enhancers. Enhancers are short DNA fragments that, when activated, signal to the cell to create certain proteins. Several exciting ML models have been developed and interpreted in an effort to learn more about the sequence features and their inter-dependencies (collectively termed syntax or grammar) that drive enhancer activity []. These include models that can be used to design cell type [] or tissue specific enhancers [], a particular exciting application with implications in synthetic biology. Despite this, defining the mechanistic roles of enhancer features during development or in tissues with complex patterns of expression remains a substantial challenge.

Very simple paradigm

Building models that capture both cis and trans gene regulation (WIP)

To take the above aim a step further, we can also ask questions about the roles that these enhancers and other regulatory elements (REs) play in coordinating what exactly a given cell does. Much work has been done to develop gene regulatory network (GRNs) models of biological systems, but linking these GRNs to the CPs, and other phenotypes, remains a challenge.

My journey

“The road goes ever on and on…”