ABSTRACT
Antimicrobial peptides are short amino acid sequences that may be antibacterial, antifungal, and antiviral. Most machine learning methodologies applied to identifying antibacterial peptides have developed feature vectors of identical lengths for each peptide in a given dataset although the peptides themselves may differ in number of amino acids. Features are often chosen which represent certain periodic patterns in the peptide sequence without any initial guidance as to whether such patterns are relevant for the classification task at hand. This can result in the construction of a large number of irrelevant features in addition to relevant features. To help alleviate these issues, we choose to extract a feature vector from individual amino acid feature representations through the application of bidirectional Long Short-Term Memory recurrent neural networks. The Long Short-Term Memory network recursively iterates along both directions of the given amino acid sequence and ultimately extracts a finite length feature vector that is then used to classify the peptide. This work demonstrates the application of Long Short-Term Memory recurrent neural networks to classification of antibacterial peptides and compares it to a Random Forest classifier and a k-nearest neighbor classifier.
Subject(s)
Neural Networks, Computer , Pore Forming Cytotoxic Proteins , Sequence Analysis, Protein/methods , Algorithms , Anti-Bacterial Agents , Computational Biology , Databases, Protein , Pore Forming Cytotoxic Proteins/chemistry , Pore Forming Cytotoxic Proteins/classification , Pore Forming Cytotoxic Proteins/geneticsABSTRACT
BACKGROUND: Cue-induced cocaine craving incubates during abstinence from cocaine self-administration. Expression of incubation ultimately depends on elevation of homomeric GluA1 alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptors in the nucleus accumbens (NAc). This adaptation requires ongoing protein translation for its maintenance. Aberrant translation is implicated in central nervous system diseases, but nothing is known about glutamatergic regulation of translation in the drug-naïve NAc or after incubation. METHODS: NAc tissue was obtained from drug-naïve rats and from rats after 1 or >40 days of abstinence from extended-access cocaine or saline self-administration. Newly translated proteins were labeled using 35S-Met/Cys or puromycin. We compared basal overall translation and its regulation by metabotropic glutamate receptor 1 (mGlu1), mGlu5, and N-methyl-D-aspartate receptors (NMDARs) in drug-naïve, saline control, and cocaine rats, and we compared GluA1 and GluA2 translation by immunoprecipitating puromycin-labeled proteins. RESULTS: In all groups, overall translation was unaltered by mGlu1 blockade (LY367385) but increased by mGlu5 blockade (MTEP). NMDAR blockade (AVP) increased overall translation in drug-naïve and saline control rats but not in cocaine/late withdrawal rats. Cocaine/late withdrawal rats exhibited greater translation of GluA1 (but not GluA2), which was not further affected by NMDAR blockade. CONCLUSIONS: Our results suggest that increased GluA1 translation contributes to the elevated homomeric GluA1 alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor levels in the NAc that mediate incubation. Additional contributions to incubation-related plasticity may result from loss of the braking influence on translation normally exerted by NMDARs. Apart from elucidating incubation-related adaptations, we found a suppressive effect of mGlu5 on NAc translation regardless of drug exposure, which is opposite to results obtained in the hippocampus and points to heterogeneity of translational regulation between brain regions.