XSEDE Science Successes
Teaching Computers to Recognize Unhealthy Guts
Teaching Computers to Recognize Unhealthy Guts
SDSC's Gordon assists in micorbiome study
Published January 18, 2017
By Tiffany Fox
A new proof-of-concept study by researchers from the University of California San Diego has succeeded in training computers to "learn" what a healthy versus an unhealthy gut microbiome looks like based on its genetic makeup. Since this can be done by genetically sequencing fecal samples, the research suggests there is great promise for new diagnostic tools that are, unlike blood draws, non-invasive.
As recent advances in scientific understanding of Parkinson's disease and cancer immunotherapy have shown, our gut microbiomes – the trillions of bacteria, viruses and other microbes that live within us – are emerging as one of the richest untapped sources of insight into human health.
The problem is these microbes live in a very dense ecology of up to one billion microbes per gram of stool. Imagine the challenge of trying to specify all the different animals and plants in a complex ecology like a rain forest or coral reef – and then imagine trying to do this in the gut microbiome, where each creature is microscopic and identified by its DNA sequence.
Determining the state of that ecology is a classic ‘Big Data' problem, where the data is provided by a powerful combination of genetic sequencing techniques and supercomputing software tools. The challenge then becomes how to mine this data to obtain new insights into the causes of diseases, as well as novel therapies to treat them.
The new paper, titled "Using Machine Learning to Identify Major Shifts in Human Gut Microbiome Protein Family Abundance in Disease," was presented last month at the IEEE International Conference on Big Data. It was written by a joint research team from UC San Diego and the J. Craig Venter Institute (JCVI). At UC San Diego, it included Mehrdad Yazdani, a machine learning and data scientist at the California Institute for Telecommunications and Information Technology's (Calit2) Qualcomm Institute; Biomedical Sciences graduate student Bryn C. Taylor and Pediatrics Postdoctoral Scholar Justine Debelius; Rob Knight, a professor in the UC San Diego School of Medicine's Pediatrics Department as well as the Computer Science and Engineering Department and director of the Center for Microbiome Innovation; and Larry Smarr, Director of Calit2 and a professor of Computer Science and Engineering. The UC San Diego team also collaborated with Weizhong Li, an associate professor at JCVI.
Metagenomics and Machine Learning
The software to carry out the study was developed by Li and run on the data-intensive Gordon supercomputer at the San Diego Supercomputer Center (SDSC), an Organized Research Unit of UC San Diego, using 180,000 core-hours. That's equivalent to running a personal computer 24 hours a day for about 20 years.
The work began with a genetic sequencing technique known as "metagenomics," which breaks up the DNA of the hundreds of species of microbes that live in the human large intestine (our "gut"). The technique was applied to 30 healthy people (using sequencing data from the National Institutes of Health's Human Microbiome Program), together with 30 samples from people suffering from the autoimmune Inflammatory Bowel Disease (IBD), including those with ulcerative colitis and with ileal or colonic Crohn's disease. This resulted in sequencing around 600 billion DNA bases, which were then fed into the Gordon supercomputer to reconstruct the relative abundance of these species; for instance, how many E. coli are present compared to other bacterial species.
Since each bacterium's genome contains thousands of genes and each gene can express a protein, this technique made it possible to translate the reconstructed DNA of the microbial community into hundreds of thousands of proteins, which are then grouped into about 10,000 protein families.
To discover the patterns hidden in this huge pile of numbers, the researchers harnessed what they refer to as "fairly out-of-the-bag" machine-learning techniques originally developed for spam filters and other data mining applications. Their goal was to use these algorithms to classify major changes in the protein families found in the gut bacteria of both healthy subjects and those with IBD, based on the DNA found in their fecal samples.
The researchers first used standard biostatistics routines to identify the 100 most statistically significant protein families that differentiate health and disease states. These 100 protein families were then used as a "training set" to build a machine learning classifier that could classify the remaining 9,900 protein families in diseased versus healthy states. The goal was to find a "signature" for which protein families were elevated or suppressed in disease versus healthy states.
The process is akin to training a computer to recognize the different flavors of fruit juices – something a human toddler could do intuitively, albeit from a limited perspective.
"From your past experiences drinking juice, you know the difference between orange, apple, and cranberry juice," Taylor noted. "Your future decision about what juice you are drinking will be based on your past preferences. But it's really hard to figure out what apple juice tastes like without experiencing it first."
They have to train the computer, in other words, to recognize what apple juice tastes like – or in this case, what a "healthy" microbiome looks like by clustering data according to bacteria.
"You can try to categorize healthy and sick people by looking at their intestinal bacterial composition," explained Taylor, "but the differences are not always clear. Instead, when we categorize by the bacterial protein family levels, we see a distinct difference between healthy and sick people. This is because proteins are the workhorses of biology, and by analyzing the proteins produced by these bacteria, we can get an idea of what the bacteria are doing in your gut."
The machine-learning approach is effective, notes Yazdani, precisely because it's statistically based. "The rules are not set in stone," he said. "What you need is past data and past experiences from patients, and then based on statistics or distribution you make your decisions. You let the data speak for itself."
Since Smarr suffers from Crohn's disease, he has been working with Knight's Center for Microbiome Innovation to advance research in this area. "Because of the exponential increase in the data on your daily changing gut microbiome, it will be essential to develop new machine-learning approaches to bring the biomedically important facets to light," said Smarr.
In the future, the researchers hope to expand their analysis, using SDSC's new Comet supercomputer, from 10,000 protein families to one million individual genes, each of which codes for a protein which can be expressed in the gut microbiome. "Scalable methods for quickly identifying such anomalies between health and disease states will be increasingly valuable for biological interpretation of sequence data," researchers wrote in the paper.
"We wanted a fast turn-around," said Yazdani. "That's really important, especially for clinical data."
Following peer review, the paper was one of only 20 percent of the 423 submissions that were accepted as regular papers for the 2016 IEEE International Conference on Big Data. The conference was held in Washington, D.C. on Dec. 5-8, 2016. Yazdani and Taylor attended and presented the paper.
As an Organized Research Unit of UC San Diego, SDSC is considered a leader in data-intensive computing and cyberinfrastructure, providing resources, services, and expertise to the national research community, including industry and academia. Cyberinfrastructure refers to an accessible, integrated network of computer-based resources and expertise, focused on accelerating scientific inquiry and discovery. SDSC supports hundreds of multidisciplinary programs spanning a wide variety of domains, from earth sciences and biology to astrophysics, bioinformatics, and health IT. SDSC's Comet joins the Center's data-intensive Gordon cluster, and are both part of the National Science Foundation's XSEDE (Extreme Science and Engineering Discovery Environment) program.
A visualization of protein families is projected onto researchers Mehrdad Yazdani and Bryn Taylor in the UC San Diego Center for Microbiome Innovation. Selfie by Bryn Taylor.
In this figure from the paper, some representative protein families are on the left-hand side of each graph. Each colored dot represents the abundance of that protein family (measured on a logarithmic scale) for a particular fecal sample taken from a healthy person (green dots) or a person with IBD (red, blue, purple dots). Selected for the top two graphs was a "training set" of 100 protein families that statistically differentiated healthy from disease states, which shows that for patients with IBD, certain protein families are either over-abundant (first graph) or under-abundant (second graph) compared with healthy subjects. The lower two graphs are the results from the machine learning algorithm, which discovered the protein families that had similar patterns in the remaining 9,900 protein families. Note that since the scale is logarithmic, these differences in abundance are often 100 to 1 or more.
- XSEDE Resources, Trinity Enable Non-Human Primate Reference Transcriptome Resource to Support Study of Genes in Our Closest Relatives
- Turtle Tree of Life
- Region 1 Champions meet at Idaho National Laboratory
- Crash test simulations expose real risks
- NSF supports development of new arctic maps
- How was the planet Earth formed?
- Exploring Large Data for Scientific Discovery
- XSEDE Value Added
- Scholars program helps realize dream
- Making sense of cyberinfrastructure
- XSEDE15 Wrap Up
- Bioinformatics Scripts Solutions
- XSEDE15 Plenary Panel
- Polymer Potential
- The Future of NSF Advanced Computing Infrastructure
- 2015 International Summer School on HPC Challenges
- A Catalyst for Complexity
- As Austin Grows So Does Its Traffic Woes
- The University of Tennessee, Knoxville, Wins Second Place in an International Student Supercomputing Competition
- PSC Receives NSF Award for Bridges Supercomputer
- Innovative New Supercomputers Increase Nation's Computational Capacity and Capability
- Exploring Competitive Balance
- A Direct Bridge
- The Dopamine Transporter
- XSEDE Supercomputers Laid the Foundation for an Unprecedented Simulation of Cosmological Evolution
- Big Data Needs Big Funding
- XSEDE helps create a more effective way to assemble genomic information
- Of Micelles and Machines
- XSEDE Allocation System to Receive Makeover
- Internet2: Advancing Science in the Age of Big Data
- XSEDE User Portal At Your Fingertips: Mobile App
- Researchers Study Air Pollution
- Dan Stanzione: New Executive Director at TACC
- People of XSEDE: Campus Champions - Preaching the HPC Gospel
- XSEDE and Blue Waters Go Supernova
- Two at a Time
- Show Him the Money
- Cosmic Slurp
- Turning Salt into the Unknown
- Looking Inside Images
- Farming the Wind
- Breaking out of the Digital Graveyard
- The Mechanism of Short-term Memory
- Open Science and Industry Collaboration
- XSEDE, Prace Call for Requests of Joint Support
- XSEDE Wins HPCWire Award
- Shields to Maximum, Mr. Scott
- The Ultimate Timekeeper
- Blue Waters, XSEDE sign collaborative agreement
- People of XSEDE - Outreach programs set XSEDE apart
- Wrangler Reels in Award
- The Great Comet: NSF awards $12 Million Grant to SDSC to deploy Comet
- Meet the Gribbles
- 2013 Nobel Prize in Chemistry winners bring HPC to the lab
- XSEDE helps create a more effective way to assemble genomic information
- XSEDE facilitates large-scale image analysis to understand diseases
- XSEDE announces new campus briding services and tools
- XSEDE, NSF Release Cloud Survey Report
- XSEDE13: Programming Competition Allows Students to "Geek Out" and Gain Crucial Skillsets
- Katlin Thaney gave XSEDE13 Keynote: Gateways for Open Science
- XSEDE13 conference selects best papers, posters visualizations and more
- XSEDE13 speaker tells how turbulence simulations help make movie magic
- XSEDE13 Plenary Talk: Accelerating Brain Research with Supercomputers
- Invited speakers announced for Extreme Scaling Workshop - Heterogenous Computing
- XSEDE13 speaker LeManuel "Lee" Bitsóí: Democratizing Scientific Research
Read more about Bitsóí's talk at this year's conference
- More than 70 students from 4 continents gain HPC skills at fourth annual Summer School
- Registration opens for Extreme Scaling Workshop 2013
- Campus Champions Fellows Named
- Campus Champions program reaches 200 members
- Rock Snot Genomics: University of Texas researchers use advanced sequencing and TACC's Ranger supercomputer to uncover origin of common algae
- Experiencing some turbulence: Researchers Take on One of Physics' Most Important and Enduring Problems
- Register now for Virtual School summer courses on data-intensive and many-core computing
- XSEDE seeks a Scientific Workflow Specialist for Extended Collaborative Support Service
Applications are due May 31, 2013
- XSEDE13 schedule now available online
- Students from high school to grad school levels invited to participate in programming contest at XSEDE13 high performance computing conference
- SDSC's Gordon enables discoveries in the study of genetics Read about Gordon's role in pinpointing the genetic patterns underlying autism-spectrum disorders, schizophrenia and similar brain conditions.
- XSEDE, National Computational Science Institute offer summer workshops for educators
- XSEDE13 Student Day applications due May 15 High school and undergraduate students get hands-on experience in computational science and interact with expert researchers
- XSEDE upgrades to Internet2's 100G Network
- XSEDE13 Registration now open!
- Get to know XSEDE Staff XSEDE Allocations Manager Ken Hackworth: The Man, The Myth, The Legend
- Two sponsors commit to XSEDE13 conference: Cray and Intel .
- Texas Unleashes Stampede
- Swirling Secrets-Understanding the turbulence of gases
- Blacklight helps researchers develop better materials for carbon capture
- Journey to the limits of spacetime
- Students invited to participate in XSEDE13 Multiple ways for high school, undergraduate, and graduate students to get involved; funding support available.
- XSEDE Call for Humanities, Arts and Social Science ProjectsIf you and your collaborators need to access to large collections of digital data, need more computer power, or require substantial storage capacity and computing power – please share it with XSEDE.
- XSEDE needs your feedback! If you received an invitation to complete the 2013 User Satisfaction Survey, please take 10 minutes today to share your comments about the XSEDE user experience.
- XSEDE deploys Globus Online for data transfer The first official software service on XSEDE has been accepted for production deployment
- The Stampede Era Begins XSEDE supercomputer now operational and available to the national open science community
- Call for ParticipationInternational Summer School on HPC Challenges in Computational Sciences
- XSEDE, European Grid Infrastructure seek collaborative use cases
Deadline extended to March 8!
- XSEDE offers free online parallel computing course Learn to use parallel computers more efficiently and productively
- NICS makes the top of Green500 list XSEDE partner recognized for energy-conscious high-performance computer, Beacon
- XSEDE's John Towns appointed to Compute Canada board of directors Board includes leaders in industry, academia, and computational research
- STILL ACCEPTING RESPONSES to Cloud Use Survey from XSEDE, NSF All researchers encouraged to respond and help shape future of cloud computing in XSEDE
- Make room for Stampede: TACC expands data center for new supercomputer
Read more about the new data center at TACC
See TACC Deputy Director, Dan Stanzione describe the new center
- SDSC welcomes Gordon supercomputer as a research powerhouse
Read more about SDSC's Gordon
- Campus Bridging Early Adopter Program issues Call For Proposals to be submitted Dec. 1-9
Read more about the program
- XSEDE12 announced -- first conference of Extreme Science and Engineering Discovery Environment
Read more about XSEDE12
- PSC, SGI Team Up on Shared-Memory Supercomputer
Read more about PSC's shared-memory supercomputer
- Pittsburgh Supercomputing Center Wins High-Performance Computing Award
Read more about PSC
- Blacklight Goes to Work at the Pittsburgh Supercomputing Center
Read more about Blacklight
- Ranger supercomputer's lifespan extended one year as part of NSF XD initiative.
Read more about Ranger
- Kraken set to deliver 2 billionth CPU hour, sustains 96 percent utilization
Read more about Kraken
- TACC Offers New, Broader Computational Biology Software Stack to Open Science Community.
Read more about biology software stack
- ACM launches new Special Interest Group on High Performance Computing. Join by Nov. 18 for special rate.
Read more about the new SIGHPC
- 'What Are You Working on Today,' Ranger, Jaguar and iForge?
Read more about TACC's Ranger supercomputer
Read more about ORNL's Jaguar supercomputer
Read more about NCSA's iForge supercomputer
- Adventures with HPC Accelerators, GPUs and Intel MIC Coprocessors
Read more about experiences with new hardware
- Developing Scientific Computing Communities
Read more about development efforts
- Indiana University to create the National Center for Genome Analysis Support, which will be integrated with XSEDE resources
Read more about the NCGAS at IU
- Scientists use XSEDE/TeraGrid resources to determine how shock waves move through solids
Read more about 'super-elastic shock waves'
- XSEDE upgrades network
Read more about the XSEDE upgrade
- Richard Tapia, Rice University mathematician and professor and member of XSEDE outreach team, receives National Medal of Science
Watch the Oct. 21 webcast
Read more about Tapia's award
Learn more about Richard Tapia
- Stampede's comprehensive capabilities to bolster U.S. open science computational resources
Read more about Stampede
Watch a video of Jay Boisseau, director of TACC, discussing Stampede
- SDSC announces scalable, high-performance data storage cloud
Read more about SDSC cloud
- Appro and SDSC Gordon supercomputer to provide up to 35M IOPS
Read more about SDSC's Gordon
- Dr. Barry Schneider from the National Science Foundation to describe XSEDE in the Oklahoma Supercomputing Symposium keynote, Oct. 11-12
Read more about Dr. Schneider's keynote
Go to symposium site
- Students research solar cells with HPC
Read more about HPC and solar research
- Seeing Is Believing: Extreme Digital visualization and data analysis resources help researchers derive insights from massive data sets
Read more about Extreme Digital
- New "Memory Advantage Program" on Blacklight at the Pittsburgh Supercomputing Center
Read more about PSC's MAP
- XSEDE project brings advanced cyberinfrastructure, digital services, and expertise to nation's scientists and engineers
Read more about XSEDE
- Watch the John Towns video
- How XSEDE will facilitate collaborative science
Read more about XSEDE and collaboration