Analyzing Genomic Contexts of Antibiotic Resistance

We annotated the hallmark proteins of class 1-4 integrons, IS elements, and conjugative systems to determine the MGE contexts other than plasmid around the ARGs. All contigs from metagenomic assemblies and RefSeq genomes were subjected to this analysis. For integrons we used integrase proteins (IntI) as the hallmark. We searched ORFs against a database containing six sequences, the representative sequences of IntI1-4 (AAQ16665.1, AAT72891.1, AAO32355.1, and 99031763) and the two outgroup sequences (P0A8P6.1 and P0A8P8.1), using diamond blastp (version 2.0.13) with –id 80 –subject-cover 80. For IS elements we used IS-associated transposases as the hallmark. We searched ORFs against the transposases retrieved from ISFinder^{60 (link)} using diamond blastp (version 2.0.13) with –id 80 –subject-cover 80. For the conjugative elements we used the hallmark proteins provided by the CONJscan database^{61 (link)}. We searched for alignments to all modules (e.g., CONJ, MOB, typeB, typeC, etc.) included in the CONJscan HMMs, using hmmsearch (version 3.2.1) with score threshold (-T) of 40. For each and every ARG ORF annotated on the genomic or metagenomic contigs, we determined distances to the closest IntI, IS transposases, and conjugative system proteins, respectively. The ARGs found within 100 Kbp of IS transposases or conjugative system proteins were assigned to be putatively associated with those MGEs, since the size of known composite transposons range up to 80 Kbp⁶² and the size of conjugative mobile elements up to 100 Kbp^{63 (link)}. The ARGs found within 10 Kbp from IntI were assigned to be associated with integrons, since the majority of integrons in bacterial genomes have 10 Kbp or shorter cassette array length^{64 (link)}.