DNA damage is constantly occurring in cells, either due to external sources or as a result of internal cellular metabolic reactions and physiological activities. Accurate repair of such DNA damages is critical to avoid mutations and chromosomal rearrangements linked to diseases including cancer, immunodeficiencies, neurodegeneration, and premature aging. A team of researchers at Massachusetts General Hospital and the National Cancer Research Centre have identified a way to repair genetic damage and prevent DNA alterations using machine learning techniques. The researchers state that it is possible to learn more about how cancer develops and how to fight it if we understand how DNA lesions originate and repair. Therefore, they hope that their discovery will help create better cancer treatments while also protecting our healthy cells. To combat challenges to DNA integrity, cells have evolved systems that detect DNA lesions and initiate a signaling cascade that promotes DNA repair, referred to as the DNA damage response (DDR).

Researchers at Tomsk Polytechnic University jointly with the University of Chemistry and Technology (Prague) conducted a series of experiments which proved that artificial neural networks can accurately identify DNA damage caused by UV radiation. In the future, this approach can be used in modern medical diagnostics. An article, dedicated to those studies, was published in the Biosensors and Bioelectronics journal. According to the authors, the ways UV could affect the DNA structure, especially with short-term irradiation, remain practically unstudied. UV radiation is also known to cause cancer.

Even if the 8-oxo-G artifacts are twice the context-specific background level (i.e., GIVG_T 2), this corresponds to only a 5 to 10% increase in the overall base-level error rate (summed over all sequence contexts; Figure 1, A to C), which is less than the intersample variability of error rates at a fixed oxoQ. A 5% increase in the base-level error rate results in a minor, if any, increase in false-positive mutation calls (Figure 1, E and F), because calling algorithms are designed to handle typical levels of sequencing error. Only at GIVG_T 5 (equivalent to oxoQ 35) do the additional errors from 8-oxo-G become comparable to the sum of all other errors and have an adverse impact on variant calling. The vast majority of samples in TCGA exhibit only minor 8-oxo-G damage that has minimal impact on mutation calling. Consequently, the claim that 73% of TCGA sequencing runs have extensive damage is misleading.

Following the Comment of Stewart et al., we repeated our analysis on sequencing runs from The Cancer Genome Atlas (TCGA) using their suggested parameters. We found signs of oxidative damage in all sequence contexts and irrespective of the sequencing date, reaffirming that DNA damage affects mutation-calling pipelines in their ability to accurately identify somatic variations. Previously, we devised a metric termed the global imbalance value (GIV) to evaluate how mutagenic damage affects sequencing accuracy (1). We showed that mutagenic damage is pervasive in public sequencing datasets and confounds the identification of somatic variants with low to moderate (1 to 5%) allelic frequency. Following our publication, the principle of global imbalance was incorporated by the International Cancer Genome Consortium (ICGC) as one of five measures used to construct a quality rating for each cancer genome (2).