First, clinical samples from patients often contain a significant amount of host DNA in addition to pathogen DNA. Sequencing these samples directly can be inefficient, as the proportion of usable pathogen data may be low. Although experimental methods such as hybridization or amplicon-based approaches effectively enrich pathogen DNA, residual host-derived DNA often remains. Therefore, additional bioinformatics-based host DNA filtering steps are essential to enhance the accuracy of pathogen sequence analyses.

Second, unlike the human genome, pathogens exhibit considerable diversity within species. The high mutation rates of RNA viruses and the frequent emergence of multiple subtypes and antibiotic resistance variants in bacteria present significant challenges for standard single-reference genome alignment methods. Moreover, when analyzing multiple pathogen species simultaneously, distinguishing between them can be difficult due to shared nucleotide sequences. To address these issues, Celemics employs a combined analytical strategy that integrates alignment-based analysis with de novo assembly. By generating longer contigs (contiguous sequences) through de novo assembly, this approach effectively distinguishes subtypes, variants, and homologous sequences across different species. This method allows for the efficient detection of unknown variants or novel subtypes and supports accurate taxonomic classification and pathogen characterization in complex mixed samples.