Study identifies common misuses of spike-in normalization

Researchers at University of California San Diego have published new guidelines that could help scientists significantly improve their results when quantifying the interactions between DNA and proteins. Understanding these interactions is critical to our understanding of human biology in general, and can also help scientists develop new treatments for a wide range of diseases, including many cancers.

The researchers focused on spike-in normalization, a widely-used molecular biology technique that helps ensure accurate and reliable results. Spike-in normalization involves adding a known quantity of chromatin (DNA and the associated proteins) to a sample before it is studied, which helps researchers account for variations between multiple samples being compared to one another. Spike-in normalization is particularly useful for comparing two conditions – such as to evaluate the impact of a drug (comparing treated vs untreated) or a mutation or a deletion in a key gene (comparing normal vs mutant).

By exploring publicly-available datasets that utilize spike-in normalization, the researchers were able to identify common scenarios where spike-in normalization is easily misused and in which building in additional quality control measures and other “guardrails” into the technique could improve results or avoid misinterpretation. By reanalyzing these data and conducting additional experiments of their own, the team was able to develop a list of nine key recommendations for researchers using spike-in normalization that could greatly increase the accuracy of their results. These measures include ensuring consistent quality control steps are taken, following best practices for computational analysis, and validating the results with other analysis techniques.

“Many studies utilize spike-in normalization, and our results call the biological conclusions drawn from this approach into question,” said senior author Alon Goren, Ph.D.

Our recommendations can help account for some of the pitfalls of spike-in normalization so we can still reap the benefits of this valuable technique.”

Alon Goren, University of California – San Diego

The study published on September 13, 2024 in Nature Biotechnology and was conducted by Lauren Patel and Yuwei Cao at UC San Diego and Eric Mendenhall, Ph.D. at HudsonAlpha Institute for Biotechnology. The study was co-led by Alon Goren, Ph.D., and Christopher Benner, Ph.D., both associate professors in the Department of Medicine at UC San Diego School of Medicine.