DEEPred: Automated Protein Function Prediction with Multi-task Feed-forward Deep Neural Networks.

Automated protein function prediction is critical for the annotation of uncharacterized protein sequences, where accurate prediction methods are still required. Recently, deep learning based methods have outperformed conventional algorithms in computer vision and natural language processing due to the prevention of overfitting and efficient training. Here, we propose DEEPred, a hierarchical stack of multi-task feed-forward deep neural networks, as a solution to Gene Ontology (GO) based protein function prediction. DEEPred was optimized through rigorous hyper-parameter tests, and benchmarked using three types of protein descriptors, training datasets with varying sizes and GO terms form different levels. Furthermore, in order to explore how training with larger but potentially noisy data would change the performance, electronically made GO annotations were also included in the training process. The overall predictive performance of DEEPred was assessed using CAFA2 and CAFA3 challenge datasets, in comparison with the state-of-the-art protein function prediction methods. Finally, we evaluated selected novel annotations produced by DEEPred with a literature-based case study considering the 'biofilm formation process' in Pseudomonas aeruginosa. This study reports that deep learning algorithms have significant potential in protein function prediction; particularly when the source data is large. The neural network architecture of DEEPred can also be applied to the prediction of the other types of ontological associations. The source code and all datasets used in this study are available at: https://github.com/cansyl/DEEPred .

The first total synthesis of the cyclodepsipeptide pipecolidepsin A.

Pipecolidepsin A is a head-to-side-chain cyclodepsipeptide isolated from the marine sponge Homophymia lamellosa. This compound shows relevant cytotoxic activity in three human tumour cell lines and has unique structural features, with an abundance of non-proteinogenic residues, including several intriguing amino acids. Although the moieties present in the structure show high synthetic difficulty, the cornerstone is constituted by the unprecedented and highly hindered γ-branched ß-hydroxy-α-amino acid D-allo-(2R,3R,4R)-2-amino-3-hydroxy-4,5-dimethylhexanoic acid (AHDMHA) residue, placed at the branching ester position and surrounded by the four demanding residues L-(2S,3S,4R)-3,4-dimethylglutamine, (2R,3R,4S)-4,7-diamino-2,3-dihydroxy-7-oxoheptanoic acid, D-allo-Thr and L-pipecolic acid. Here we describe the first total synthesis of a D-allo-AHDMHA-containing peptide, pipecolidepsin A, thus allowing chemical structure validation of the natural product and providing a robust synthetic strategy to access other members of the relevant head-to-side-chain family in a straightforward manner.

Use of Gene Ontology Annotation to understand the peroxisome proteome in humans.

The Gene Ontology (GO) is the de facto standard for the functional description of gene products, providing a consistent, information-rich terminology applicable across species and information repositories. The UniProt Consortium uses both manual and automatic GO annotation approaches to curate UniProt Knowledgebase (UniProtKB) entries. The selection of a protein set prioritized for manual annotation has implications for the characteristics of the information provided to users working in a specific field or interested in particular pathways or processes. In this article, we describe an organelle-focused, manual curation initiative targeting proteins from the human peroxisome. We discuss the steps taken to define the peroxisome proteome and the challenges encountered in defining the boundaries of this protein set. We illustrate with the use of examples how GO annotations now capture cell and tissue type information and the advantages that such an annotation approach provides to users. Database URL: http://www.ebi.ac.uk/GOA/ and http://www.uniprot.org.

UniProtJAPI: a remote API for accessing UniProt data.

Programmatic access to the UniProt Knowledgebase (UniProtKB) is essential for many bioinformatics applications dealing with protein data. We have created a Java library named UniProtJAPI, which facilitates the integration of UniProt data into Java-based software applications. The library supports queries and similarity searches that return UniProtKB entries in the form of Java objects. These objects contain functional annotations or sequence information associated with a UniProt entry. Here, we briefly describe the UniProtJAPI and demonstrate its usage.