An increasingly common theme in biological research is the need to manage enormous amount of information in a way that provides integrated understanding of living systems. This is seen in various “omics” approaches, including genomics, metabolomics, proteomics, and so on, that focus on genes, metabolites, and proteins, respectively. As technological advances uncover overwhelming numbers of these factors, scientists claim that this greatly challenges their effort to explain, prevent, and change biological and disease outcomes of interest.
In the context of this magnitude problem in biology, various solutions have been proposed that allow for a type of “dimensionality reduction” or "causal selection"–essentially reducing a large set of factors down to those that matter the most. This proposal provides a novel framework for performing this causal selection in the context of causal explanation. This framework involves classifying different types of causes that are common in biology and clarifying the principled reasons that guide using these differences to sort important causal factors from all others.
Clarifying these differences and their role in causal selection is important for many reasons. These selection methods can (i) identify the most valuable targets for controlling biological outcomes (including disease prevention and treatment), (ii) locate factors with the most explanatory power, and (iii) prevent reductive tendencies in biology that over-include irrelevant lower-level details.
This project provides an analysis of these causal selection procedures in a way that relies on rigorous analytic philosophy work on causation and careful examination of biological case-studies, including the scientific methods, goals, and reasoning they involve.
