I'm Yaan

A

Work on computational biology with a minor in biophysics, develop computational tools for protein folding & engineering and now work on statistical methods to address the molecular mechanisms of lipid exchange at membrane contact sites and interpret how the genetic variations impact the lipid transport and biosynthesis applications.

About Me

Bioinformaticist & Data Analyst

Known as : N. J. Cheung
Phone : +44-(0)75 123 13219

Expertise : Bioinformatics
E-mail : yaan.jang[AT]gmail.com

Work involves molecular mechanisms of lipid exchange at membrane contact sites, lipid transporters, protein engineering, folding pathways & structure prediction, GPCR-G protein binding & signalling, data analysis & modeling, sequencing data analysis, machine learning, Bayesian statistics

Contact Me

Coding Lines

Projects Done

Talks

Visiting

What I'm interested in

I am also interested in the application of machine learning and optimization algorithms to protein design, folding and its tertiary structure prediction. These include Boltzmann machine, deep neural networks, Bayesian optimization algorithms, and swarm intelligence. I am developing SENET (a sequentially evolving framework) that can be used in structural informatics and to aid molecular design

Deep Learning & Optimization

Broadly, I'm interested in developing computational methods that can be leveraged in design of biological molecules for practical applications. Especially, I work on designing proteins that bind with high affinity & specificiy to target mollecules and are used for therapeutics, creating catalysts for chemical reactions via enzyme design, and developing sampling algorithms & forcefields for validating the designs

Protein Folding & Design

I leverage Amo and SATAY (in silico) to study the molecular mechanisms of lipid exchange at membrane contact sites and how the transporters devliver the lipids between mitochondria and ER, as well as other organelles. Being rooted in physical principles, identifying evolutionary signatures allows for residue-level interpretation of their roles in lipid transportation and inter-molecular interactions

Lipid Transport & Data Anlysis

Projects selected

Amoai

Amo (Amoai v2.7.5) is elegant with modular codes, efficient with high-performance, precise at high accuracy, and fast at high speed

SENET

SENET is a sequentially evolving framework and projected to build a learning engine for recognizing our life in a digital way

Sibe

Sibe is a powerful biological engine and used for protein tertiary structure prediction, deep learning, statistical analysis, and optimization

MoDyFing

MoDyFing is an approach for predicting protein folding pathways and tertiary structure from its primary sequence

iTooU

An integrated approach for iteratively predicting protein folding pathways and tertiary structure from its primary sequence

OptiFel

OptiFel is a data-driven modeling method based on a heterogeneous particle swarm optimization (CHPSO) algorithm

My Resume

Appointment & Education

Bioinformatics scientist In Biochemistry

10/2019 - present University of Oxford

Molecular mechanisms of lipid traffic at contact sites

Sequencing data analysis for SATAY

Research Associate in Physics

07/2018 - 09/2019 University of Cambridge

Machine learning and Bayesian inference

The sequentially evolving neural network framework

Postdoctoral Researcher in Cognitive Science

11/2016 - 05/2018 DGIST

Structural bioinformatics & fMRI data analysis

Developed Sibe for protein, brain and cognitive studies

Ph.D. in Bioinformatics & Biophysics (NDVS)

2012 - 2016

2014-2016 The University of Chicago

2012-2014 Shanghai Jiao Tong University

Publications

^*corresponding author; ^† equal contribution

2021

S. van Schie, A. T. J. Peter, N. J. Cheung, B. Kornmann. Rewiring phospholipid biosynthesis reveals robustness in membrane homeostasis and uncovers lipid regulatory players. Submitted 2021.
N. J. Cheung^*, A. T. J. Peter, B. Kornmann. Leri: a web-server for identifying proteinfunctional networks from evolutionary couplings. Computational and Structural Biotechnology Journal 19, 2021: 3556-3563. [Webserver]
Y.-C. Xu, T.-J. ShangGuan, X.-M. Ding^*, N. J. Cheung^*. Accurate prediction of protein torsion angles using evolutionary signatures and recurrent neural network, 2021.

2020

N. J. Cheung. Residue communities reveal evolutionary signatures of protein function in Britain STEM, Dec 2020.
N. J. Cheung. SATAY project on next-generation sequencing to functionally screen the genome of Saccharomyces cerevisiae, 2020. Web-server under SATAY Project.

2019

N. J. Cheung. A general evolving framework for deep neural networks. (SENET project).
N. J. Cheung. Protein design from evolutionary couplings among amino acids using Leri. Web-server under Amoai Project.
Team@Amoai. De novo sequence assembly of a human genome with ultra-long reads from Nanopore sequencing using Amoai.
N. J. Cheung. Evolutionary couplings recognize barcodes between G_α and GPCRs. Web-server under Amoai Project.
N. J. Cheung, C. Willacy. Convolutional neural network for identifying seismic events from vertical seismic profile data. Shell (internal use only), 2019
N. J. Cheung, W. Yu. Sibe: a computation tool to apply protein sequence statistics to predict folding and design in silico. BMC bioinformatics 20(1), 2019: 1-11.

2018

Team@Amoai. Sibe as part of Amo is projected to build a learning engine for recognizing our life from the next generation sequencing data.
N. J. Cheung. NiNet: Learning better by a self-renewal network. It is in developing based on Phsior of Sibe but will not be maintained.
N. J. Cheung, W. Yu. De novo protein structure prediction using ultra fast molecular dynamics simulation. PLoS ONE 13(11) 2018: e0205819.

2017

N. J. Cheung. Folding proteins from amino acid sequences using computational intelligence algorithm. Attempt to answer "How far can computational intelligence go in protein folding?"
N. J. Cheung. Sibe: C++ library of biological engine. Completed in studies on protein science and is extending to analysis on sequencing data.
N. J. Cheung, X.-M. Ding, H.-B. Shen. A non-homogeneous cuckoo search algorithm based-on quantum mechanism for real-parameter optimization. IEEE Transactions on Cybernetics 47(2): 391-402, Feb. 2017. [ PDF] [bibtex]. [Supplementary Material]

2016

N. J. Cheung, W. Yu, K. F. Freed, T. R. Sosnick. Statistical analysis on protein evolutionary information for stability and design. Poster @DGIF in South Korea, Dec. 2016. DOI: 10.5281/zenodo.1172905. [ PDF]
N. J. Cheung, W. Yu, T. R. Sosnick, K. F. Freed. Computational investigation on WW-domain proteins specified by evolutionary information.
N. J. Cheung. Protein folding pathways, structure prediction and design by computational intelligence. Doctoral Thesis.
N. J. Cheung, X.-M. Ding, H.-B. Shen. A non-homogeneous firefly algorithm and its convergence analysis. Journal of Optimization Theory and Applications, 170(2): 616-628, Jan. 2016. [Supplementary Material]
N. J. Cheung, X.-M. Ding, H.-B. Shen. Protein folds recognized by an intelligent predictor based-on evolutionary and structural information. Journal of Computational Chemistry, 37(4): 426-436, 2016. [ PDF] [bibtex]