This article presents causal analysis approaches implemented in Ingenuity Pathway Analysis (IPA) for interpreting gene-expression data. The methods leverage a large-scale causal network derived from the Ingenuity Knowledge Base, which contains over 40,000 nodes representing genes, chemicals, and biological processes, and over 1.48 million edges representing experimentally observed cause-effect relationships. The approach uses statistical methods to infer upstream regulators and downstream effects, and to identify causal relationships relevant to the observed expression changes. Four key algorithms are described: Upstream Regulator Analysis (URA), Mechanistic Networks (MN), Causal Network Analysis (CNA), and Downstream Effects Analysis (DEA). URA identifies upstream regulators that may explain observed expression changes, while MN builds on URA to identify causal mechanisms. CNA extends URA to include paths with multiple regulators, and DEA predicts downstream biological functions and diseases. The methods use two main scores: an overlap P-value, which measures the enrichment of regulated genes in the dataset, and a Z-score, which assesses the match between observed and predicted up/down regulation patterns. The Z-score is particularly useful for predicting the activation state of a regulator. The algorithms are applied to real-world datasets, including gene-expression data from breast cancer cells treated with beta-estradiol and endothelial cells stimulated with TNF. The results show that the methods can accurately identify upstream regulators and predict downstream effects, demonstrating their utility in understanding the biological mechanisms underlying gene-expression data. The approach is validated by comparing results with known biological knowledge and showing consistency across different cell types and species. The methods are implemented in IPA and are available for use in analyzing gene-expression datasets.This article presents causal analysis approaches implemented in Ingenuity Pathway Analysis (IPA) for interpreting gene-expression data. The methods leverage a large-scale causal network derived from the Ingenuity Knowledge Base, which contains over 40,000 nodes representing genes, chemicals, and biological processes, and over 1.48 million edges representing experimentally observed cause-effect relationships. The approach uses statistical methods to infer upstream regulators and downstream effects, and to identify causal relationships relevant to the observed expression changes. Four key algorithms are described: Upstream Regulator Analysis (URA), Mechanistic Networks (MN), Causal Network Analysis (CNA), and Downstream Effects Analysis (DEA). URA identifies upstream regulators that may explain observed expression changes, while MN builds on URA to identify causal mechanisms. CNA extends URA to include paths with multiple regulators, and DEA predicts downstream biological functions and diseases. The methods use two main scores: an overlap P-value, which measures the enrichment of regulated genes in the dataset, and a Z-score, which assesses the match between observed and predicted up/down regulation patterns. The Z-score is particularly useful for predicting the activation state of a regulator. The algorithms are applied to real-world datasets, including gene-expression data from breast cancer cells treated with beta-estradiol and endothelial cells stimulated with TNF. The results show that the methods can accurately identify upstream regulators and predict downstream effects, demonstrating their utility in understanding the biological mechanisms underlying gene-expression data. The approach is validated by comparing results with known biological knowledge and showing consistency across different cell types and species. The methods are implemented in IPA and are available for use in analyzing gene-expression datasets.