Me
Shang-Hung Shih
Software Engineer @ Taiwan AILabs
shshih8497@gmail.com
Fast Learner
I am passionate with applying new technology to interesting projects, and like to participate in challenging projects.

Biography

Shang-Hung Shih has worked as a software engineer at the genomics team in Taiwan AILabs for two years. Prior to this, He graduated from National Taiwan University (NTU) and Academia Sinica with master degree in Bioinformatics, and worked with Prof. Chien-Yu Chen.

Timeline

Jan. 2020 -
Software Engineer @ Taiwan AI Labs
TAIGenomics, PubmedKB, QCheck web app development
Mar. 2019 - Dec. 2019
Software Engineer Intern @ Taiwan AI Labs
Variant Annotator
Supervisor and Mentor: Stefan Hong
Jul. 2018 - Dec. 2020
M.S. @ NTU c4Lab
2018 - 2021
M.S. degree @ NTU and Academia Sinica
Genome and Systems Biology Degree Program
Jan. 2015 - Aug. 2018
Research Assistant @ CSMU bio608
Dockerized Next-Generation-Sequencing Analysis
2014 - 2018
B.Sc. degree @ CSMU
Biomedical Science

Selected Projects

Moneyfire Landing Page and Mobile App

Develop a mobile application to help users track and manage asset growth.

Typescript(React Native, Next.js) / Firebase / AWS(Amplify)
PubmedKB Web App

Working on a PubmedKB web app for entities relation visualization of NLP.

Typescript(ReactJS) / Node.js / Neo4j / Docker / Kubernetes
QCheck Web App

Built proteomics data visualization for QCheck web app.

Javascript(Vue.js) / php(laravel) / MySQL / AWS
TAIGenomics Web App

Developed signup system, variant table, IGV viewer, and pedigree editor for TAIGenomics web app to help genetic analysis and diagnosis.

Typescript(ReactJS) / Node.js / MongoDB / Docker / Kubernetes
V-score: Pathogenic Score Prediction

Developed rule-based ACMG, ML-based v-score and inheritance pattern interpretation system for variant prioritization.

Python(ML) / Docker
Variant Annotator

Designed and built a variant annotator with multi-processing and MySQL to solve the inefficiency of traditional annotation tools for matching 4 million variants to databases with 9 billion data, reduced the annotation time by 58%, and shortened the time for a whole genome sample from 2 hours to 50 minutes.

Python / MySQL / Docker
1st place in PIXNET 6th Hackathon Final: Travel Tech (Travel Charger)

Used Pixnet articles for correlation analysis and IG post data for sentiment analysis to provide personalized attraction recommendation and trip planning. I was mainly responsible for building relation graph for attraction recommendation.

Python / Neo4j / MySQL / Docker
Constructed Transcription Factor Binding Profiles using Deep Learning

Predicting the sequence specificities of DNA-binding proteins by convolutional neural network.

Python(DL) / MySQL / Docker
ngs-main-wes: Dockerized Multi-threads Somatic Paired WES Analysis Pipeline

For general varaints calling (GATK3 Mutect2 & GATK4 Mutect2 with PONs) and annotation (Phial based on Oncotator, paraSNP based on Annovar).

Python / Docker
Top