Isaac Ho Email & Phone Number

Redwood City, CA

• Ported flagship analysis pipeline to AWS to augment surge throughput capacity (EC2, ParallelCluster, Docker)
• Designed and implemented novel hashing/locking scheme to reduce contention for multithreaded detection of genomic contamination (C++, Python)
• Designed and implemented a front-to-end solution to enable custom regression testing on core pipeline results, including browser-based navigation of result tables (Python, Javascript, jQuery)

• Lead engineer on 3-man team for Meraculous, a short-read genome assembler that handles gigabase-scale genomes on high­ performance clusters using a deBruijn graph approach. In the Assemblathon 2, Meraculous was.
• Applied for and received $24k AWS-in-Education grant. Ported to AWS and performed advanced tuning to drive cost for de-novo human assembly to under 30 hours / $70 on single EC2 instance
• Extended the Landau-Vishkin read alignment algorithm to support local alignment and a mismatch penalty while maintaining big-O runtime bound
• Designed and implemented a multithreading model for the core modules of Meraculous, using 65k google:sparse_hashes and spinlocks designated by DNA prefix to balance memory footprint and efficiency
• Rewrote core modules in C++, using efficient binary-to-text conversion, custom DNA sequence class, and careful profiling to achieve >5x throughput over existing Perl version
• Developed novel lightweight hash data structure for k­mer lookups that does not need to store the keys themselves, storing the entire assembly k­mer graph in approximately 4 bytes per genomic base ( an improvement of.

• Bioniks's core product was a MATLAB prototype that enhanced the tracking of skeletal markers provided by a single Microsoft Kinect for use in orthopedic evaluations, for several key clinical metrics, it was shown to.
• Designed and built a real-time 24fps Java port using OpenCV/OpenNI
• Added a GUI to visualize key metrics like knee flexion and hip drop
• Project delivered in just 80 hours, 50% under budget, despite no previous domain experience

• Developed the GO-Elite plugin for Cytoscape, providing researchers with an integratedway to mine gene expression data using overrepresentation analysis.
• Integration involved manipulation of core Cytoscape data structures, implementation of an OPAL webservice client, and using threading to boost the Swing GUI performance
• Implemented RESTful webservice for BridgeDb, providing easy and robust programmatic id­mapping across various standard biological id systems

• Lead engineer on 5-man team for developing overlap-layout-consensus assembly tools for Sanger data, converting prototype code to industry-grade C++ code with unit tests and best practices
• Built structural variant analysis tool in Perl to bulk-detect insertions/deletions/transversions in Arabidopsis paired-end reads in support of original research
• Core developer for JAZZ Sanger-read assembler

Bachelor'S Degree, Computer Science With Honors

Education record