A5016330639- RNA and protein synthesis mechanisms
- Multimodal Machine Learning Applications
- Advanced Graph Neural Networks
- RNA modifications and cancer
- Topic Modeling
- RNA Interference and Gene Delivery
- Advanced Image and Video Retrieval Techniques
- Semantic Web and Ontologies
- Domain Adaptation and Few-Shot Learning
- RNA Research and Splicing
- DNA and Nucleic Acid Chemistry
- Advanced biosensing and bioanalysis techniques
- DNA and Biological Computing
- Bacterial Genetics and Biotechnology
Oregon State University
2023-2024
Nanjing University
2019-2020
Abstract Motivation RNA design is the search for a sequence or set of sequences that will fold to desired structure, also known as inverse problem folding. However, designed by existing algorithms often suffer from low ensemble stability, which worsens long design. Additionally, many methods only small number satisfying MFE criterion can be found each run These drawbacks limit their use cases. Results We propose an innovative optimization paradigm, SAMFEO, optimizes objectives (equilibrium...
<title>Abstract</title> RNA design aims to find an sequence that can fold into a given target structure, which enables the creation of artificial molecules with specific function, and has numerous applications in medicine. Computationally, it is particularly challenging due two levels combinatorial explosion: exponentially large space many competing structures for each design. As result, heuristic methods such as local search have been popular this task, but they cannot keep up explosion. We...
Relevance search over a knowledge graph (KG) has gained much research attention. Given query entity in KG, the problem is to find its most relevant entities. However, relevance function hidden and dynamic. Different users for different queries may consider from angles of semantics. The ambiguity more noticeable presence thousands types entities relations schema-rich which challenged effectiveness scalability existing methods. To meet challenge, our approach called RelSUE requests user...
Relevance search is to find top-ranked entities in a knowledge graph (KG) that are relevant query entity. ambiguous, particularly over schema-rich KG like DBpedia which supports wide range of different semantics relevance based on numerous types relations and attributes. As users may lack the expertise formalize desired semantics, supervised methods have emerged learn hidden user-defined from user-provided examples. Along this line, paper we propose novel generative model KGs for search,...
The tasks of designing RNAs are discrete optimization problems, and several versions these problems NP-hard. As an alternative to commonly used local search methods, we formulate as continuous develop a general framework for this based on generalization classical partition function which call "expected function". basic idea is start with distribution over all possible candidate sequences, extend the objective from sequence distribution. We then use gradient descent-based methods improve...
Motivation: RNA design aims to find at least one sequence that folds with the highest probability into a designated target structure, but some structures are undesignable in sense no them. Identifying is useful delineating and understanding limit of designability, has received little attention until recently. In addition, existing methods on undesignability not scalable interpretable. Results: We introduce novel graph representation new general algorithmic framework efficiently identify...
This paper proposes Multi-modAl Retrieval model via Visual modulE pLugin (MARVEL), which learns an embedding space for queries and multi-modal documents to conduct retrieval. MARVEL encodes with a unified encoder model, helps alleviate the modality gap between images texts. Specifically, we enable image understanding ability of well-trained dense retriever, T5-ANCE, by incorporating visual module's encoded features as its inputs. To facilitate retrieval tasks, build ClueWeb22-MM dataset...
RNA design is the search for a sequence or set of sequences that will fold into predefined structures, also known as inverse problem folding. While numerous methods have been invented to find capable folding target structure, little attention has given identification undesignable structures according minimum free energy (MFE) criterion under Turner model. In this paper, we address gap by first introducing mathematical theorems outlining sufficient conditions recognizing then proposing...
Relevance search is to find top-ranked entities in a knowledge graph (KG) that are relevant query entity. ambiguous, particularly over schema-rich KG like DBpedia which supports wide range of different semantics relevance based on numerous types relations and attributes. As users may lack the expertise formalize desired semantics, supervised methods have emerged learn hidden user-defined from user-provided examples. Along this line, paper we propose novel generative model KGs for search,...