1887

Abstract

Deep learning (DL) is a subset of Artificial Intelligence employing neural networks that require the use of a training set and are modelled on circuit pathways in the human brain. Whilst AI is a burgeoning field today, its roots are in the 1950s. DL algorithms use multiple layers to progressively extract higher level features from the raw input. Many different architectures of neural network exist, and for the most part are involved in applications with image recognition. Some recent uses of DL include self-driving cars; there are also creative projects using DL to create fake faces or human poses for use as models, composing music, creating novel art from different styles of art and writing fake news.

spp. are well-known as important producers of bioactive compounds such as antibiotics. These bioactive compounds are often encoded as secondary metabolites in the organisms by large gene clusters such as non-ribosomal peptide synthases (NRPS) and polyketide synthases (PKS). The NRPS and PKS assemble peptides using enzymatic units arranged in modules that can function in an iterative or sequential fashion independent of messenger RNA. Each NRPS or PKS is capable of assembling one type of peptide. Fusions of the two also exist.

Here we have trained deep learning neural networks to provide us with simulated “fake” secondary metabolite sequences. We examine the characteristics of these sequences and how they could be used to guide us with antibiotic discovery.

Loading

Article metrics loading...

/content/journal/acmi/10.1099/acmi.ac2020.po0511
2020-07-10
2021-03-02
Loading full text...

Full text loading...

http://instance.metastore.ingenta.com/content/journal/acmi/10.1099/acmi.ac2020.po0511
Loading

Most cited this month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error