“Li has significantly improved the bioinformatics infrastructure and expertise needed to analyzed invivo ChIP-Seq and RNA-Seq data, acquired from neuropsychiatric models of addiction and depression here at MSSM. He is extraordinarily competent, highly efficient, and continually provides insightful consultation on ways to refine and interpret the vast quantities of data produced by next generation sequencing.”
About
* Extensive experience developing machine learning algorithms for bioinformatics
*…
Activity
-
today I announce the publication of "DeepRegFinder: Deep Learning-Based Regulatory Elements Finder". We compared our method with other established…
today I announce the publication of "DeepRegFinder: Deep Learning-Based Regulatory Elements Finder". We compared our method with other established…
Shared by Li Shen
-
human body is naturally symmetrical and that provides a sort of weak labels. our new work leverages this cue for pretraining based on Siamese…
human body is naturally symmetrical and that provides a sort of weak labels. our new work leverages this cue for pretraining based on Siamese…
Shared by Li Shen
-
Filter 12 is distinctly associated with the active enhancer class and you can see it clearly has a peak detector for H3K27ac. Use DeepRegFinder…
Filter 12 is distinctly associated with the active enhancer class and you can see it clearly has a peak detector for H3K27ac. Use DeepRegFinder…
Shared by Li Shen
Experience
-
-
-
-
-
-
-
-
-
-
-
-
-
Education
-
Nanyang Technological University
-
Activities and Societies: President of Xinyu photography club, Toastmasters club
• Thesis about machine learning techniques for microarray cancer classification.
• Completed thesis in 3.5 years; conferred degree in 02/2007. -
-
• Fudan University Deans List, 1998-2001
Publications
-
End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design
arXiv
See publicationWe develop an end-to-end training algorithm for whole-image breast cancer diagnosis based on mammograms. It has the advantage of training a deep learning model without relying on cancer lesion annotations. Our approach is implemented using an all convolutional design that is simple yet provides superior performance in comparison with the previous methods. With modest model averaging, our best models achieve an AUC score of 0.91 on the DDSM data and 0.96 on the INbreast data. We also demonstrate…
We develop an end-to-end training algorithm for whole-image breast cancer diagnosis based on mammograms. It has the advantage of training a deep learning model without relying on cancer lesion annotations. Our approach is implemented using an all convolutional design that is simple yet provides superior performance in comparison with the previous methods. With modest model averaging, our best models achieve an AUC score of 0.91 on the DDSM data and 0.96 on the INbreast data. We also demonstrate that a trained model can be easily transferred from one database to another with different color profiles using only a small amount of training data.
-
Automatic genome segmentation with HMM-ANN hybrid models
Workshops on Machine Learning in Computational Biology at the Annual Conference on Neural Information Processing Systems (NIPS)
See publicationUnsupervised learning that combines hidden Markov models with artificial neural networks to perform genome segmentation based on large-scale ChIP-seq data.
-
β-catenin mediates stress resilience through Dicer1/microRNA regulation
Nature
Our findings establish β-catenin as a critical regulator in the development of behavioural resilience, activating a network that includes Dicer1 and downstream microRNAs. We thus present a foundation for the development of novel therapeutic targets to promote stress resilience.
Other authors -
-
Chronic cocaine-regulated epigenomic changes in mouse nucleus accumbens
Genome Biology
Large-scale sequencing analysis project that reveals epigenomic changes in cocaine mice.
Other authors -
-
STAR: an integrated solution to management and visualization of sequencing data
Bioinformatics
A genome browser for viewing epigenomic data. My prototype and last project at UCSD, which has grown into a fully featured, production ready website by the hard work of Dr. Tao Wang.
Other authors -
-
diffReps: Detecting Differential Chromatin Modification Sites from ChIP-seq Data with Biological Replicates
PLOS ONE
ChIP-seq differential analysis package, featuring automated analysis pipeline and integrated annotation.
Other authors -
-
GBNet: Deciphering regulatory rules in the co-regulated genes using a Gibbs sampler enhanced Bayesian network approach
BMC Bioinformatics
Bayesian model selection with simulated annealing for regulatory motif rule learning.
Other authors -
-
Reducing multiclass cancer classification to binary by output coding and SVM
Computational Biology and Chemistry
See publicationUsing error-correction coding for multiclass classification problems.
-
Dimension Reduction-Based Penalized Logistic Regression for Cancer Classification Using Microarray Data
IEEE/ACM Transactions on Computational Biology and Bioinformatics
See publicationMachine learning for cancer classification with high-dimensional microarray data.
Projects
-
Shenlab@Github
Shenlab's unofficial (but most updated) website: our software projects, publications, etc.
Other creators -
-
ngs.plot: Quick Mining and Visualization of Next-Generation Sequencing Data by Integrating Genomic Databases
- Present
An integrated solution to quickly visualize large amount of next-generation sequencing data featuring both command line and web interfaces. It has become very popular among bioinformatic researchers.
Other creators -
-
diffReps: automated ChIP-seq differential analysis
- Present
An automated pipeline to analyze large ChIP-seq data featuring robust background estimation, negative binomial test and parallel processing.
Other creators -
-
STAR genome browser
-
I initiated, designed and implemented the prototype of STAR genome browser – an integrated solution for management and visualization of next-generation sequencing data using PHP, MySQL and JavaScript on Apache server. It is adopted by the acclaimed NIH roadmap epigenomics project. Manuscript accepted by Bioinformatics.
Other creators -
Recommendations received
4 people have recommended Li
Join now to viewMore activity by Li
-
Just finished the 1st virtual FBI retreat on Zoom. Thank you Zoom for developing such a wonderful tool! Perfect video experience.
Just finished the 1st virtual FBI retreat on Zoom. Thank you Zoom for developing such a wonderful tool! Perfect video experience.
Posted by Li Shen
-
Our amazing NewYork-Presbyterian nurses, physicians and other health care team members have been tirelessly and selflessly fighting this battle…
Our amazing NewYork-Presbyterian nurses, physicians and other health care team members have been tirelessly and selflessly fighting this battle…
Liked by Li Shen
-
"Scientific Reports published more than 19, 871 papers in 2019, and so a position in the top 100 most downloaded articles is an extraordinary…
"Scientific Reports published more than 19, 871 papers in 2019, and so a position in the top 100 most downloaded articles is an extraordinary…
Posted by Li Shen
-
I was just notified that our paper (https://lnkd.in/d7Sv5TH) about breast cancer detection using deep learning is a top 100 downloaded articles 2019…
I was just notified that our paper (https://lnkd.in/d7Sv5TH) about breast cancer detection using deep learning is a top 100 downloaded articles 2019…
Shared by Li Shen
-
Proud to be part of this international effort to fight breast cancer with mammography & AI: https://lnkd.in/dUsuYdK
Proud to be part of this international effort to fight breast cancer with mammography & AI: https://lnkd.in/dUsuYdK
Shared by Li Shen
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Li Shen in United States
127 others named Li Shen in United States are on LinkedIn
See others named Li Shen