The paper introduces a method for combining independent sources of evidence using p-values, specifically applied to sequence homology searches. The authors propose calculating the product of p-values from multiple motifs to estimate the likelihood of a sequence belonging to a specific family. They derive a formula and algorithm (QFAST) to compute the statistical distribution of the product of n independent p-values. This method effectively combines information from multiple motifs, leading to more accurate and sensitive sequence homology searches. The MAST algorithm, which incorporates this method, is available for interactive use and downloading. The paper also discusses the implementation and validation of the method, demonstrating its accuracy and effectiveness in classifying sequences into families.The paper introduces a method for combining independent sources of evidence using p-values, specifically applied to sequence homology searches. The authors propose calculating the product of p-values from multiple motifs to estimate the likelihood of a sequence belonging to a specific family. They derive a formula and algorithm (QFAST) to compute the statistical distribution of the product of n independent p-values. This method effectively combines information from multiple motifs, leading to more accurate and sensitive sequence homology searches. The MAST algorithm, which incorporates this method, is available for interactive use and downloading. The paper also discusses the implementation and validation of the method, demonstrating its accuracy and effectiveness in classifying sequences into families.