How Digitalization and Algorithms Are Changing the Knowledge in Biomedicine

Algorithms are essential to the modern transformation of biomedicine, enabling the analysis of large and complex biological datasets through computational models, machine learning, and network analysis. These indirect methods allow for quick hypothesis generation, discovering potential interactions, and gaining insights into biological systems that would be difficult to achieve with experimental methods alone. However, relying only on computational predictions can lead to misinterpretation and false positives. Therefore, we must combine algorithms with rigorous experimental validation to ensure scientific accuracy and reliability. Improvements in algorithm efficiency, data integration, and validation strategies continually strengthen their robustness, making them vital tools in systems biology. Combining algorithms with experimentation creates a dynamic cycle of hypothesis testing and refinement, leading to deeper insights into biological functions and advancements in diagnostics and therapeutics. We illustrate, in the text, what happens when preventive checks on data reliability are missing. Using STRING and similar resources, we show how to apply Popper's Falsification Principle to interactomics, for falsification, not mere exploration.