ActTRANS: Functional classification in active transport proteins based on transfer learning and contextual representations

Taju, Semmy Wellem; Shah, Syed Muazzam Ali; Ou, Yu‐Yen

doi:10.1016/j.compbiolchem.2021.107537

Cited by 7 publications

(3 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The use of pre‐trained language models in protein sequences has shown great potential in various bioinformatics tasks. Recently, researchers have started applying these models to biological sequence analysis [21–25]. By training these models on large amounts of sequence data, researchers can extract informative embeddings that capture the patterns and relationships between amino acid residues in protein sequences.…”

Section: Recent Advances In Features Generationmentioning

confidence: 99%

“…These models can learn complex relationships between words and sentences in a given language and generate highly informative embeddings that capture the semantics and syntax of the language. Recently, researchers have started applying these models to biological sequence analysis [21][22][23][24][25]. By training these models on large amounts of sequence data, researchers can extract informative embeddings that capture the patterns and relationships between amino acid residues in protein sequences.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Recent advances in features generation for membrane protein sequences: From multiple sequence alignment to pre‐trained language models

Ou,

Ho,

Chang

2023

Proteomics

View full text Add to dashboard Cite

Membrane proteins play a crucial role in various cellular processes and are essential components of cell membranes. Computational methods have emerged as a powerful tool for studying membrane proteins due to their complex structures and properties that make them difficult to analyze experimentally. Traditional features for protein sequence analysis based on amino acid types, composition, and pair composition have limitations in capturing higher‐order sequence patterns. Recently, multiple sequence alignment (MSA) and pre‐trained language models (PLMs) have been used to generate features from protein sequences. However, the significant computational resources required for MSA‐based features generation can be a major bottleneck for many applications. Several methods and tools have been developed to accelerate the generation of MSAs and reduce their computational cost, including heuristics and approximate algorithms. Additionally, the use of PLMs such as BERT has shown great potential in generating informative embeddings for protein sequence analysis. In this review, we provide an overview of traditional and more recent methods for generating features from protein sequences, with a particular focus on MSAs and PLMs. We highlight the advantages and limitations of these approaches and discuss the methods and tools developed to address the computational challenges associated with features generation. Overall, the advancements in computational methods and tools provide a promising avenue for gaining deeper insights into the function and properties of membrane proteins, which can have significant implications in drug discovery and personalized medicine.

show abstract

Section: Recent Advances In Features Generationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Recent advances in features generation for membrane protein sequences: From multiple sequence alignment to pre‐trained language models

Ou,

Ho,

Chang

2023

Proteomics

View full text Add to dashboard Cite

show abstract

“…The model has been widely used for its excellent performance in text classification [91]. When dealing with domain-specific classification tasks such as pharmaceutical technology, it is required to construct small domain sample datasets and pre-train the model [17,92,93].…”

Section: Bidirectional Encoder Representations For Transformers With ...mentioning

confidence: 99%

Data-Driven Technology Roadmaps to Identify Potential Technology Opportunities for Hyperuricemia Drugs

et al. 2022

View full text Add to dashboard Cite

Hyperuricemia is a metabolic disease with an increasing incidence in recent years. It is critical to identify potential technology opportunities for hyperuricemia drugs to assist drug innovation. A technology roadmap (TRM) can efficiently integrate data analysis tools to track recent technology trends and identify potential technology opportunities. Therefore, this paper proposes a systematic data-driven TRM approach to identify potential technology opportunities for hyperuricemia drugs. This data-driven TRM includes the following three aspects: layer mapping, content mapping and opportunity finding. First we deal with layer mapping.. The BERT model is used to map the collected literature, patents and commercial hyperuricemia drugs data into the technology layer and market layer in TRM. The SAO model is then used to analyze the semantics of technology and market layer for hyperuricemia drugs. We then deal with content mapping. The BTM model is used to identify the core SAO component topics of hyperuricemia in technology and market dimensions. Finally, we consider opportunity finding. The link prediction model is used to identify potential technological opportunities for hyperuricemia drugs. This data-driven TRM effectively identifies potential technology opportunities for hyperuricemia drugs and suggests pathways to realize these opportunities. The results indicate that resurrecting the pseudogene of human uric acid oxidase and reducing the toxicity of small molecule drugs will be potential opportunities for hyperuricemia drugs. Based on the identified potential opportunities, comparing the DNA sequences from different sources and discovering the critical amino acid site that affects enzyme activity will be helpful in realizing these opportunities. Therefore, this research provides an attractive option analysis technology opportunity for hyperuricemia drugs.

show abstract

Identification of efflux proteins based on contextual representations with deep bidirectional transformer encoders

Taju

Shah

2021

Analytical Biochemistry

View full text Add to dashboard Cite

ActTRANS: Functional classification in active transport proteins based on transfer learning and contextual representations

Cited by 7 publications

References 37 publications

Recent advances in features generation for membrane protein sequences: From multiple sequence alignment to pre‐trained language models

Recent advances in features generation for membrane protein sequences: From multiple sequence alignment to pre‐trained language models

Data-Driven Technology Roadmaps to Identify Potential Technology Opportunities for Hyperuricemia Drugs

Identification of efflux proteins based on contextual representations with deep bidirectional transformer encoders

Contact Info

Product

Resources

About