Abstract: We describe a method and system for detecting pathogens in a sample, particularly an air sample. The method is designed to run in real-time alongside sequencing processing of the sample so that results are quickly available to a user. Each input sequence read in the plurality of sequence reads is compared using a first classification algorithm and assigned to a taxon. An input sequence is classified as potentially pathogenic when the taxon to which the input sequence read is assigned corresponds to a pathogen taxon. Each input sequence read classified as potentially pathogenic is compared using a second classification algorithm to a plurality of target sequences in a second database which is selected based on the pathogen taxon. A confidence score is used as part of the decision-making process to decide if an individual read represents a genuine match to a pathogen.