Identificación de bacteriocinas con propiedades neuropeptídicas empleando un modelo de aprendizaje automático no supervisado

Bookmark (0)
Please login to bookmark Close

The gut microbiota has been found to interact with the brain through the microbiota-gut-brain axis, according to numerous studies, as it is highlighted in Ullah et al., 2023. This pathway has gained relevance in recent years due to increasing evidence suggesting that the gut microbiota hold a strong impact on neurodevelopmental disorders such as autism or ADHD, among others (Iannone et al., 2019).
According to this, different strategies are being developed to unveil the actual role of these microorganisms and the molecular mechanisms surrounding these diseases. Some of the approaches are related to the application of antibiotics, fecal microbiota transplantation (Borrego-Ruiz y Borrego, 2024), brain imaging or microbiome sequencing (Buxbaum et al., 2012). However, we focused our attention on bioinformatics, an underexplored field with high potential in Biotechnology. In particular, we concentrated our efforts on the combined use of deep learning and language models like BERT to create a contextual predictor that allows us to infer peptide function based on its sequence. This approach not only helps us understand how microbial interactions influence host physiology but also enables us to investigate the potential role of bacteriocins (proteins of interest) as neurotransmitters, providing new insights into their involvement in neural communication and signaling pathways.
In this context, peptide-based therapy has become a potential new method for disease treatment due to the important role it plays in biological routes. Compared with the previously mentioned methods, peptides have fewer side effects and present some highly desired qualities like high specificity, low production cost, low toxicity, and easy synthesis. However, despite all the advantages, they are still very difficult and time-consuming to produce. That is the reason why computer-aided protein analysis could help us enlighten the mechanisms behind these types of syndromes while reducing costs and making it more accessible.

​The gut microbiota has been found to interact with the brain through the microbiota-gut-brain axis, according to numerous studies, as it is highlighted in Ullah et al., 2023. This pathway has gained relevance in recent years due to increasing evidence suggesting that the gut microbiota hold a strong impact on neurodevelopmental disorders such as autism or ADHD, among others (Iannone et al., 2019).
According to this, different strategies are being developed to unveil the actual role of these microorganisms and the molecular mechanisms surrounding these diseases. Some of the approaches are related to the application of antibiotics, fecal microbiota transplantation (Borrego-Ruiz y Borrego, 2024), brain imaging or microbiome sequencing (Buxbaum et al., 2012). However, we focused our attention on bioinformatics, an underexplored field with high potential in Biotechnology. In particular, we concentrated our efforts on the combined use of deep learning and language models like BERT to create a contextual predictor that allows us to infer peptide function based on its sequence. This approach not only helps us understand how microbial interactions influence host physiology but also enables us to investigate the potential role of bacteriocins (proteins of interest) as neurotransmitters, providing new insights into their involvement in neural communication and signaling pathways.
In this context, peptide-based therapy has become a potential new method for disease treatment due to the important role it plays in biological routes. Compared with the previously mentioned methods, peptides have fewer side effects and present some highly desired qualities like high specificity, low production cost, low toxicity, and easy synthesis. However, despite all the advantages, they are still very difficult and time-consuming to produce. That is the reason why computer-aided protein analysis could help us enlighten the mechanisms behind these types of syndromes while reducing costs and making it more accessible. Read More