Even More Open Source Medicine and Machine Learning, in This Week’s Wrap Up

Open Source Wrap Up: November 14-20, 2015

Group of Biohackers Start Work on Open Source Insulin

More than 370 million people suffer from diabetes worldwide, and these people rely on regular insulin injections to regulate the amount of sugar in their blood. Despite this major need for insulin, there is no generic version available on the market, and the high cost results in it having limited availability in poorer parts of the world. A group of citizen and academic researchers and biohackers, led by Counter Culture Labs, have launched a project to develop a simple method for producing insulin and release the process to the public domain. They have launched a crowdfunding campaign (that has already exceeded their goal) to fund the first stage of this research.  For stage 1, ” the team will insert an optimized DNA sequence for insulin into E. coli bacteria, induce the bacteria to express insulin precursors, and verify that human proinsulin has been produced.”

Netflix Launches Open Source Platform for Continuous Delivery

Netflix has launched Spinnaker, a new open source project for continuous delivery, in conjunction with Google, Microsoft, and Pivotal. It is designed to serve as a replacement to Asgard, the model Netflix previously used to manage Amazon Web Services. Spinnaker expands on Asgard by providing support for multiple platforms, including Google Cloud Platform, and Cloud Foundry, with support for Microsoft Azure reportedly in development. Spinnaker also provides cluster management and visibility into the footprint of applications on cloud infrastructure.

Check out the code on GitHub.

Linux Goes to the International Space Station

The United Space Alliance, a NASA contractor, has decided to migrate to Linux for operations on the International Space Station (ISS). The company’s Laptops and Network Integration team is in charge of the station’s OpsLAN: a network of laptops that provide the crew with vital capabilities for day-to-day operations. The team decided to migrate from Windows to Linux in order to improve stability and reliability, and provide better control over the operating system they deploy. The Linux Foundation provided the training needed for this team to develop a solution based on Linux through classes that introduced the team members to Linux and application development on the platform.

Read More from the Linux Foundation.

Microsoft Releases Distributed Machine Learning Toolkit

Following in the wake of Google’s release of TensorFlow, Microsoft has answered with their own open source, distributed machine learning toolkit. DMTK includes a flexible framework that supports a unified interface for data parallelization, big model storage, model scheduling, and automatic pipelining. It also includes distributed algorithms for word embedding and multi-sense word embedding as well as LightLDA: a fast and lightweight system for large-scale topic modeling. DMTK offers an API designed to reduce the barrier to entry of distributed machine learning with the goal being for researchers and developers to focus more on the data, model, and training the core machine logics. By open sourcing this software, Microsoft hopes to encourage academia and the industry to get involved in the platform’s development.

The code can be found on GitHub.

Other News

Author: Ben Lloyd Pearson

Ben is an open source technologist experienced in a wide array of modern IT tools. He's done a bit of everything, from web development, to systems administration, web publishing, and beyond.