Back

Deciphering the co-evolutionary dynamics of L2 β-lactamases via Deep learning

Zhu, Y.; Gu, J.; Zhao, Z.; Chan, A. W. E.; Mojica, M. F.; Hujer, A. M.; Bonomo, R. A.; Haider, S.

2024-01-14 microbiology
10.1101/2024.01.14.575584 bioRxiv
Show abstract

L2 {beta}-lactamases, a serine-based class A {beta}-lactamases expressed by Stenotrophomonas maltophilia plays a pivotal role in antimicrobial resistance. However, limited studies have been conducted on these important enzymes. To understand the co-evolutionary dynamics of L2 {beta}-lactamase, innovative computational methodologies, including adaptive sampling molecular dynamics simulations, and deep learning methods (convolutional variational autoencoders and BindSiteS-CNN) explored conformational changes and correlations within the L2 {beta}-lactamase family together with other representative class A enzymes including SME-1 and KPC-2. This work also investigated the potential role of hydrophobic nodes and binding site residues in facilitating the functional mechanisms. The convergence of analytical approaches utilized in this effort yielded comprehensive insights into the dynamic behaviour of the {beta}-lactamases, specifically from an evolutionary standpoint. In addition, this analysis presents a promising approach for understanding how the class A {beta}-lactamases evolve in response to environmental pressure and establishes a theoretical foundation for forthcoming endeavours in drug development aimed at combating antimicrobial resistance. SynopsisDeep learning is used to reveal the dynamic co-evolutionary patterns of L2 {beta}-lactamases. O_LIAnalysis of hydrophobic nodes and binding site residues provides a detailed understanding of both local and global dynamic evolution, which explain the functional divergences. C_LIO_LIThe employment of two distinct deep learning models, the Convolutional Variational Autoencoder (CVAE) and BindSiteS-CNN, facilitates the investigation of conformational shifts, thereby depicting the dynamic evolution of L2 {beta}-lactamases. C_LIO_LIThe effectiveness of CVAE and BindSiteS-CNN in dynamic classification is corroborated with selected features. C_LI O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=96 SRC="FIGDIR/small/575584v1_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@1db15forg.highwire.dtl.DTLVardef@168043eorg.highwire.dtl.DTLVardef@1eebb96org.highwire.dtl.DTLVardef@5d8019_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.