When scientists study gene activity using RNA sequencing, they often encounter missing data points, where certain genes appear to have no measurable expression. Understanding why this happens is important because missing values can skew analyses, leading to incorrect conclusions about how genes behave in health and disease. A recent study led by researchers from Baylor College of Medicine addresses this very challenge, investigating the factors that contribute to missing gene expression values in RNA sequencing data.

The researchers analyzed RNA sequencing data from 66 lung adenocarcinoma tumors and the surrounding normal lung tissues, searching for patterns in missingness. They found that genes with very low expression levels were more likely to be missing, which was expected because genes that are barely active are difficult to detect. Interestingly, they also found that some genes with very high expression were missing in certain samples, suggesting a phenomenon they call “true biological missingness” or TBM. These are genes that are highly active in some individuals but completely inactive in others, reflecting real biological differences rather than technical errors.

The study also explored how smoking affects missingness patterns. They discovered that some genes sensitive to tobacco smoke were more likely to exhibit missing values, supporting the idea that true biological differences between individuals can drive missingness in RNA sequencing datasets. The researchers caution that standard methods for filling in missing data, known as imputation, can introduce bias if TBM genes are included. Instead, they recommend identifying genes with true biological missingness and analyzing them separately to maintain the accuracy of gene expression studies.

Overall, this research highlights the complexity behind missing values in RNA sequencing and provides guidance for scientists to carefully handle missing data, especially when inter-individual differences may explain the absence of certain gene expressions. By understanding these nuances, researchers can improve the reliability of RNA sequencing analyses and gain clearer insights into how genes behave across different individuals and conditions.