A new algorithm to solve memory problems in large-scale human brain simulations

 Researchers have come closer towards advancing technology to create computer simulations of the brain networks using exascale-class supercomputers. Their findings are published in journal  Frontiers in Neuroinformatics.

Computer brain simulations
Credit: Forschungszentrum Jülich

The human brain is a complex network composed of approximately 100 billion neurons. With current computing power, it is impossible to simulate 100 percent working brain. Currently, researchers use simulating software called NEST to simulate the brain. NEST, a free, open-source simulation code in widespread use by the neuroscientific community and a core simulator of the European Human Brain Project.

“Since 2014, our software can simulate about one percent of the neurons in the human brain with all their connections,” says Markus Diesmann, Director at the Jülich Institute of Neuroscience and Medicine (INM-6). To achieve this impressive feat, the software requires the entire main memory of petascale supercomputers.

With NEST, the behavior of each neuron in the network is represented by a handful of mathematical equations. Future exascale computers, such as the post-K computer planned in Kobe and JUWELS in Jülich, will exceed the performance of today’s high-end supercomputers by 10- to 100-fold. For the first time, researchers will have the computer power available to simulate neuronal networks on the scale of the human brain.

While current simulation technology enabled researchers to begin studying large neuronal networks, it also represented a dead end on the way to exascale technology. Supercomputers are composed of about 100,000 small computers, called nodes, each equipped with many processors doing the actual calculations.

“Before a neuronal network simulation can take place, neurons and their connections need to be created virtually, which means that they need to be instantiated in the memory of the nodes. During the simulation, a neuron does not know on which of the nodes it has target neurons. Therefore, its short electric pulses need to be sent to all nodes. Each node then checks which of all these electric pulses are relevant for the virtual neurons that exist on this node,” explains Susanne Kunkel of KTH Royal Institute of Technology in Stockholm.

The current algorithm for network creation is efficient because all nodes construct their particular part of the network at the same time. However, sending all electric pulses to all nodes is not suitable for simulations on exascale systems.

“Checking the relevance of each electric pulse efficiently requires one Bit of information per processor for every neuron in the whole network. For a network of 1 billion neurons, a large part of the memory in each node is consumed by just this single Bit of information per neuron,” adds Markus Diesmann.

This is the main problem when simulating even larger networks: the amount of computer memory required per processor for the extra Bits per neuron increases with the size of the neuronal network. At the scale of the human brain, this would require the memory available to each processor to be 100 times larger than in today’s supercomputers. This, however, is unlikely to be the case in the next generation of supercomputers. The number of processors per compute node will increase, but the memory per processor and the number of compute nodes will rather stay the same.

The breakthrough published in Frontiers in Neuroinformatics is a new way of constructing the neuronal network in the supercomputer. Due to the algorithms, the memory required on each node no longer increases with network size. At the beginning of the simulation, the new technology allows the nodes to exchange information about who needs to send neuronal activity data to whom. Once this knowledge is available, the exchange of neuronal activity data between nodes can be organized such that a node only receives the information it requires. An additional Bit for each neuron in the network is no longer necessary.

While testing their new ideas, the scientists made an additional key insight, reports Susanne Kunkel: “When analyzing the new algorithms we realized that our novel technology would not only enable simulations on exascale systems, but it would also make simulations faster on presently available supercomputers.”

In fact, as the memory consumption is now under control, the speed of simulations becomes the main focus of further technological developments. For example, a large simulation of 0.52 billion neurons connected by 5.8 trillion synapses running on the supercomputer JUQUEEN in Jülich previously required 28.5 minutes to compute one second of biological time. With the improved data structures simulation, the time is reduced to 5.2 minutes.

“With the new technology we can exploit the increased parallelism of modern microprocessors a lot better than previously, which will become even more important in exascale computers,” remarks Jakob Jordan, lead author of the study, from Forschungszentrum Jülich.

“The combination of exascale hardware and appropriate software brings investigations of fundamental aspects of brain function, like plasticity and learning unfolding over minutes of biological time within our reach,” adds Markus Diesmann.

With one of the next releases of the simulation software NEST, the researchers will make their achievement freely available to the community as open source.

“We have been using NEST for simulating the complex dynamics of the basal ganglia circuits in health and Parkinson’s disease on the K computer. We are excited to hear the news about the new generation of NEST, which will allow us to run whole-brain-scale simulations on the post-K computer to clarify the neural mechanisms of motor control and mental functions,” says Kenji Doya of Okinawa Institute of Science and Technology (OIST).

“The study is a wonderful example of the international collaboration in the endeavor to construct exascale computers. It is important that we have applications ready that can use these precious machines from the first day they are available,” concludes Mitsuhisa Sato of the RIKEN Advanced Institute for Computer Science in Kobe.

Citation: Jordan, Jakob, Tammo Ippen, Moritz Helias, Itaru Kitayama, Mitsuhisa Sato, Jun Igarashi, Markus Diesmann, and Susanne Kunkel. “Extremely Scalable Spiking Neuronal Network Simulation Code: From Laptops to Exascale Computers.” Frontiers in Neuroinformatics 12 (2018). doi:10.3389/fninf.2018.00002.

Research funding: Helmholtz Portfolio Supercomputing and Modeling for the Human Brain (SMHB), Helmholtz young investigator group, EU 7th Framework Programme (Human Brain Project), EU Horizon 2020 research and innovation programme (Human Brain Project).

Adapted from press release by Frontiers.

Computational study finds new genetic targets for autism

Researchers at the University of Missouri created a new computational method that has connected several target genes to autism. Recent discoveries could lead to screening tools for young children and could help doctors determine correct interventions when diagnosing autism.

Unlocking the genetic causes of autism requires data-intensive computations. In 2014, the National Science Foundation (NSF) awarded $1 million in two grants to MU to install a supercomputer enabling data-intensive research and education in the fields of bioinformatics and data-driven engineering applications.

“In this study we started with more than 2,591 families who had only one child with autism and neither the parents nor the siblings had been diagnosed with autism,” said Chi-Ren Shyu, director of the Informatics Institute and the Paul K. and Dianne Shumaker Endowed Professor in the Department of Electrical Engineering and Computer Science in the MU College of Engineering. “This created a genetically diverse group composed of an estimated 10 million genetic variants. We narrowed it down to the 30,000 most promising variants, then used preset algorithms and the big data capabilities of our high-performance computing equipment at MU to ‘mine’ those genetic variables.”

The genetic samples were obtained from the Simons Foundation Autism Research Initiative. Samples from children with diagnosed cases of autism, and their unaffected parents and siblings were collected leading to more than 11,500 individuals. Using advanced computational techniques, Shyu and his team were able to identify 286 genes that were then collected into 12 subgroups that exhibited commonly seen characteristics of children on the spectrum. Of these genes, 193 potentially new genes not found in previous autism studies were discovered.

“Autism is heterogeneous, meaning that the genetic causes are varied and complex,” said Judith Miles, professor emerita of child health-genetics in the MU Thompson Center for Autism and Neurodevelopmental Disorders. “This complexity makes it tough for geneticists to get at the root of what triggers the development of autism in more conventional ways. The methods developed by Dr. Shyu and the results our team identified are giving geneticists a wealth of targets we’d not considered before–by narrowing down the genetic markers, we may be able to develop clinical programs and methods that can help diagnose and treat the disease. These results are a quantum leap forward in the study of the genetic causes of autism.”

Citation: Spencer, Matt, Nicole Takahashi, Sounak Chakraborty, Judith Miles, and Chi-Ren Shyu. “Heritable genotype contrast mining reveals novel gene associations specific to autism subgroups.” Journal of Biomedical Informatics 77 (2018): 50-61. doi:10.1016/j.jbi.2017.11.016.

Funding:National Institutes of Health, National Science Foundation, Shumaker Endowment for Biomedical Informatics, Simons Foundation.

Adapted from press release by University of Missouri.

Finding new uses for old medication using computer program DrugPredict

Researchers at Case Western Reserve University School of Medicine have developed a computer program called DrugPredict to discover new indications for old drugs. This program matches existing data about FDA-approved drugs to diseases, and predicts potential drug efficacy.

In a recent study published in Oncogene, the researchers successfully translated DrugPredict results into the laboratory, and showed common pain medications non-steroidal anti-inflammatory drugs, also known as NSAIDs could have applications for epithelial ovarian cancer.

DrugPredict was developed by co-first author QuanQiu Wang of ThinTek, LLC, and co-senior author Rong Xu, PhD, associate professor of biomedical informatics in the department of population and quantitative health sciences at Case Western Reserve University School of Medicine. The program works by connecting computer-generated drug profiles including mechanisms of action, clinical efficacy, and side effects with information about how a molecule may interact with human proteins in specific diseases, such as ovarian cancer.

DrugPredict searches databases of FDA-approved drugs, chemicals, and other naturally occurring compounds. It finds compounds with characteristics related to a disease-fighting mechanism. These include observable characteristics (phenotypes) and genetic factors that may influence drug efficacy. Researchers can collaborate with Xu to input a disease into DrugPredict and receive an output list of drugs or potential drugs with molecular features that correlate with strategies to fight the disease.

In the Oncogene study, DrugPredict produced a prioritized list of 6,996 chemicals with potential to treat epithelial ovarian cancer. At the top of the list were 15 drugs already FDA-approved to treat the cancer, helping to validate the DrugPredict approach. Of other FDA-approved medications on the list, NSAIDs ranked significantly higher than other drug classes. The researchers combined the DrugPredict results with anecdotal evidence about NSAIDs and cancer before confirming DrugPredict results in their laboratory experiments.

Citation: Nagaraj, A. B., Q. Q. Wang, P. Joseph, C. Zheng, Y. Chen, O. Kovalenko, S. Singh, A. Armstrong, K. Resnick, K. Zanotti, S. Waggoner, R. Xu, and A. Difeo. “Using a novel computational drug-repositioning approach (DrugPredict) to rapidly identify potent drug candidates for cancer treatment.” Oncogene, 2017.

DOI: 10.1038/onc.2017.328

Funding: Gynecological Cancer Translation Research Program, Case Comprehensive Cancer Center, The Mary Kay Foundation, NIH/Eunice Kennedy Shriver National Institute Of Child Health & Human Development.

Adapted from press release by Case Western Reserve University.

ResistoMap developed to track world wide microbial drug resistance

Scientists from the Federal Research and Clinical Center of Physical-Chemical Medicine, the Moscow Institute of Physics and Technology, and Data Laboratory have created the ResistoMap, an interactive visualization of gut resistome. Gut resistome is human gut microbiota potential to resist antibiotics and includes the set of all antibiotic resistance genes in the genomes of human gut microbes. Their ResistoMap will help identify national trends in antibiotic use and control antibiotic resistance on the global scale. This research is published in journal Bioinformatics.

Resistomap - interactive world map of human gut resistome
Resistomap –  an interactive world map of human gut microbiota potential to resist antibiotics.
Credit: Bioinformatics

Microbial drug resistance is caused by the extensive uncontrolled use of antibiotics in medicine and agriculture. It has been predicted that by 2050 around 10 million people will die annually due to reasons associated with drug resistance.

The ResistoMap has two main interactive work fields: a geographic map and a heat map. A user can choose the antibiotic group or country of interest to be displayed on the heat map and obtain a resistome cross section. The data can be filtered by the country of origin, gender, age, and diagnosis. The current version of the interactive map developed by the authors draws on a dataset that includes over 1600 individuals from 12 studies covering 15 countries. However, the dataset can be expanded by additional input from users reflecting the findings of new published studies in a unified format.

Using the ResistoMap, researchers fee that it is possible to estimate the global variation of the resistance to different groups of antibiotics and explore the associations between specific drugs and clinical factors or other metadata. For example, the Danish gut metagenomes tend to demonstrate the lowest resistome among the European groups, whereas the French samples have the highest levels, particularly of the fluoroquinolones, a group of broad-spectrum anti-bacterial drugs. This is in agreement with the fact that France has the highest total antibiotic use across Western Europe, while the use of antimicrobial drugs in Denmark and Germany is moderate, both in health care and agriculture. At the opposite end of the spectrum, Chinese and Russian populations appear to have increased levels of resistome, which is likely due to looser regulation policies, frequent prescription of broad-spectrum antibiotics, and their over-the-counter availability without prescription. The lowest levels of microbiota resistome are observed in the native population of Venezuela who have no documented contacts with populations of the developed countries. ResistoMap-informed analysis reveals certain novel trends that await further interpretation from the clinical standpoint.

Konstantin Yarygin, one of the creators of the visualization tool, says, “We anticipate that the exploratory analysis of global gut resistome enabled by the ResistoMap will provide new insights into how the use of antibiotics in medicine and agriculture could be optimized.”

Citation: Yarygin, Konstantin S., Boris A. Kovarsky, Tatyana S. Bibikova, Damir S. Melnikov, Alexander V. Tyakht, and Dmitry G. Alexeev. “ResistoMap—online visualization of human gut microbiota antibiotic resistome.” Bioinformatics, 2017.
doi:10.1093/bioinformatics/btx134.
Research funding: Russian Scientific Foundation.
Adapted from press release by the Moscow Institute of Physics and Technology.

Zika virus lineage and evolution

The Zika virus remains a mystery. Isolated from macaque monkeys in the Ziika Forest in Uganda in 1947, the virus was shown to infect humans not long after, but it was identified as a benign disease, with mild symptoms. For this reason, it was not heavily studied until almost 70 years later when it appeared to be associated with an unusual cluster of cases of microcephalic birth defects and Guillain-Barré Syndrome (GBS) paralysis in Brazil in 2015 and 2016.

Zika Virus. Credit: Wikipedia

If the at-least-70-year-old virus is responsible for the recently reported neurological diseases, why were the first serious effects not noticed until recently? And, why were these effects first in Brazil, very distant from its continent of apparent origin, Africa? The mysterious history of the virus matters because its details might tell us the backstory of how it came to be what it is where it is and from that, why it is doing so much damage.

But, how do you know the history of an invisible virus, which leaves no physical record? It is especially hard to know the history of Zika because the seemingly benign disease has been under-the-radar for most of its known time in human hosts.

This is where genetics can help, since single-strand RNA viruses like Zika tend to change rapidly over time and, with bioinformatics, researchers can deduce what the ancestral relationships are between different viruses collected at different places in different times from different hosts. While the first noted occurrence of the virus was in Africa, it was detected only a few years later in Asia, and separate lineages of the disease are known from both areas – a clue that the history hidden in the genes may be complicated.

“But sequence data on Zika is limited,” notes University of North Carolina at Charlotte Bioinformatics and Genomics Professor Daniel Janies. “People have made the assumption that it came out of Africa because that’s where it was discovered. However, it has not been easy to reconstruct the history of Zika with the data we have,” he said.

Janies heads a team of researchers who have recently completed a phylogenetic and geographic analysis of the available collection of Zika’s genetic sequences. The analysis provides the most complete study of the virus’s history to date and reveals specific genetic changes that occurred as the virus crossed the Pacific Ocean on its way to the Americas. An analysis of the genes involved also suggests new hypotheses to explain the virus’s association with microcephaly and GBS.

A report by Janies, Adriano de Bernardi Schneider, Jun-tao Guo, Gregorio Linchangco, Zachary Witter, Dylan Vinesett and Lambodhar Damodaran from the department of Bioinformatics and Genomics at UNC Charlotte, Robert Malone from Atheric Pharmaceutical, and Jane Homan from IoGenetics LLC appears in the issue of Cladistics.

“Our results indicate that Zika may have deep ancestry in Asia that has been under-recorded,” Janies said. “For example, not all the recent global outbreaks of Zika appear to result from a simple linear chronology of travel from the most recent past outbreak.”

“Recently there has been an outbreak of Zika in Singapore in parallel to the one in the Americas. We have updated our analyses and the Singapore Zika virus is distantly related to the viral lineage in the Americas. This lends support for the hypothesis that there are yet-to-be-discovered reservoirs of Zika virus in Asia,” Janies said.

The Cladistics report traces Zika’s phylogenetic tree through analysis of genetic sequences, combining it with the chronology and geographic information from the samples, and allows the researchers to detail the virus’ probable historical path as well as specific genetic and structural changes in the virus as it traveled to the Americas.

The researchers noted, in particular, some new mutations that began appearing in the virus as it traveled from island to island across the Pacific. Not long after these mutations appear, there are records in French Polynesia of an increase in both microcephaly and GBS. The specific nature of the new mutations in the virus also suggest to the team some possible relationships between viral infection and the severe symptoms associated with the virus in the Americas.

“We looked at the viral changes that correspond to the first reports of microcephaly and we saw the origins of these changes in the Pacific lineages,” Janies noted. “There are mutations that occurred in the part of the viral genome that codes the viral envelope protein and the ends of the viral genome that are called ‘untranslated regions.’ We focused on the envelope protein because that’s the part responsible for the entry of the virus to host’s cells. We studied the untranslated regions since they mediate the types of tissues the virus attacks and viral replication.”

Both sets of mutations suggested potential relationships to the virus’s new association with neurological and developmental problems in adults and infants.

“Members of our team found that Zika has recently started making its envelope proteins with features, called epitopes, that are similar to human proteins, which could cause a human host immune response to the virus to be diluted,” Janies said. “The theory underlying this idea is called ‘epitope mimicry.’ The similarity is advantageous to the virus because it confuses the host’s immune system and blunts the immune reaction to the virus.”

However, the researchers suspect that the human proteins being mimicked may be significant for reasons besides providing immune system “cover” for the attacking virus.

An important element of the envelope protein mutation, Janies points out, is not only in the mimicry itself, but also, in the specific genes being mimicked: “Our team members found that two of the human proteins that Zika is mimicking are involved in the signaling that goes on when the sensory organs are being formed in the fetus. These genes are called ‘Neuron Navigator Protein 2’ and ‘Human Neurogenic Differentiation Factor 4’, ” he said.

“Because these are the proteins are being mimicked, a hypothesis is that the developmental pathways that rely on the proteins may be being disrupted by the immune system,” Janies said.

The other mutations, on the untranslated regions, suggest other possible effects that might change where Zika virus infects in the body.

Although epitope mimicry hypothesis helps clarify the protein-immune interaction, the mutations in the untranslated regions may explain the types of tissues Zika attacks” UNC Charlotte Bioinformatics and Genomics graduate student Adriano de Bernardi Schneider said. “The presence of specific binding regions on untranslated regions of the Zika viral genome, called “Musashi Binding Elements” provides bases for the study of changes in tissue preference of the virus.”

In this part of the study, the authors evaluated the changes in the virus’ Musashi Binding Elements and found that they increased the efficiency of the Zika virus that is circulating in the Americas in hijacking human cells.

Musashi is a family of RNA-binding proteins in the host cells that control gene expression and the development of stem cells. The finding that Zika has mutated to be better at binding to human Musashi proteins, leads to the hypothesis that Zika is adapting to be more efficient at attacking human cells. Moreover, the role of Musashi proteins in stem cells provides another possible target for the study of developmental defects in the fetus associated with Zika infection in pregnancy.

Both the autoimmune effect and changes in the virus’ tissue specificity are working hypotheses suggested by computational models and will require further study to verify.

In contrast, the information gained from studying Zika’s phylogenetic history is of immediate importance to medicine and public health response, as this work puts the mutations in specific time and place context, at a time when the virus has nearly circled the planet, changing from place to place in its travels and leaving different variants. Many versions of the virus currently exist globally and these variants have different capabilities and effects.

“We’re tracing the lineages and the geographic links in a very rigorous way and pulling it all together, pinpointing Zika’s molecular changes in time and space – showing what actually is going on in different places,” Janies said. “Why does it matter? Well, when Zika arrives someplace is it going to be benign or dangerous? It has been both — it depends on where it is coming from.”

Citation: Adriano de Bernardi Schneider, Robert W. Malone, Jun-Tao Guo, Jane Homan, Gregorio Linchangco, Zachary L. Witter, Dylan Vinesett, Lambodhar Damodaran and Daniel A. Janies. “Molecular evolution of Zika virus as it crossed the Pacific to the Americas”. Cladistics 2016.
DOI: 10.1111/cla.12178
Research funding: Defense Advanced Research Projects Agency.
Adapted from press release by the University of North Carolina at Charlotte. 

Novel bio-signal measuring electrodes to advance health diagnosis using internet of things devices

Daegu Gyeongbuk Institute of Science and Technology (DGIST) announced that Professor Kyung-in Jang’s research team from the Department of Robotics Engineering succeeded in developing bio-signal measuring electrodes that can be mounted on Internet of Things (IoT) devices through joint research with a research team led by professor John Rogers of the University of Illinois, USA.

Optical image of bio-signal measurement electrode
design developed by Professor Jang’s research team.
The electrode generates such a large force that it holds
the circular magnet located under the glass only by
attraction (gravitation) of the magnetic field.
Credit: DGIST
The bio-signal measuring electrodes developed by the research team can be easily mounted onInternet of Things (IoT) devices for health diagnosis, thus they can measure bio-signals such as brain waves and electrocardiograms without additional analysis and measurement equipment while not interfering or restricting human activities.
Conventional hydro-gel based electrodes required external analysis and measurement devices to measure bio-signals due to their pulpy gel forms, which made their attachment to and detachment from IoT devices instable. In addition, since these electrodes were wet-bonded to the skin, there have been disadvantages that the characteristics of the electrodes deteriorated or their performance decreased when the electrodes were dried in the air over a long period.
In contrast, the electrodes developed by Professor Kyung-in Jang can be easily interlocked as if they are a part of Internet of Things (IoT) devices for health diagnosis. Also, since they are composed only of polymer and metal materials, they have the advantage of there being no possibility of drying in the air.

The bio-signal measurement electrodes developed by the research team consist of a composite material in which a magnetic material is folded with a soft and adhesive polymer, with a conductive electrode material wrapped around the composite material. The conductive electrode material electrically connects the bottom surface touching the skin and the top surface touching the electrode of the Internet of Things (IoT) device.

Electrodes with this structure reacting to the magnetic field can be easily attached and detached by using the attraction that occurs between the magnet and the electrode mounted on the IoT devices. Then, through the conductive electrode materials that connect the skin and the electrode part of the IoT device, the electric signals generated on the skin can be directly transmitted to the IoT device for health diagnosis.

The research team succeeded in storing and analyzing brain waves (electroencephalogram, EEG), electrocardiograms (ECG), eye movements (electrooculogram, EOG), and limb movements and muscle contractions (electromyogram, EMG) of the wearer for a long period through an experiment in which IoT devices with the electrodes are attached to various parts of the human body.

The bio-signal measurement electrodes can measure the bioelectric signal generated from the skin without loss or noise by using the Internet of Things (IoT) platform, thus they are expected to be applicable to the medical and healthcare fields since they cannot only measure the electrical signals of the body, but also analyze various forms of bio-signals such as body temperature change, skin change, and in-body ion concentration change.

Professor Kyung-in Jang said, “We have secured the source technology that can diagnose the state of human health anytime and anywhere by combining bio-electrode technology with Internet of Things (IoT) platforms utilizing advanced high-tech composite materials. We will carry out subsequent research to make it applicable for diseases that require ongoing medical diagnosis such as diabetes, insomnia, and epilepsy, and to make it available to people in medically vulnerable areas such as remote mountainous and rural areas.”

Citation: “Ferromagnetic, Folded Electrode Composite as a Soft Interface to the Skin for Long-Term Electrophysiological Recording”. Kyung-In Jang, Han Na Jung, Jung Woo Lee, Sheng Xu, Yu Hao Liu, Yinji Ma, Jae-Woong Jeong, Young Min Song, Jeonghyun Kim, Bong Hoon Kim, Anthony Banks, Jean Won Kwak, Yiyuan Yang, Dawei Shi, Zijun Wei, Xue Feng, Ungyu Paik, Yonggang Huang, Roozbeh Ghaffari, John A. Rogers.Advanced Functional Materials 2016 vol: 26 (40) pp: 7281-7290.
DOI: 10.1002/adfm.201603146
Adapted from press release by Daegu Gyeongbuk Institute of Science and Technology (DGIST).

Computer based model InFlo predicts cell signals and network activity

A multi-institution academic-industrial partnership of researchers led by Case Western Reserve University School of Medicine has developed a new method to broadly assess cell communication networks and identify disease-specific network anomalies. The computer-based method, called InFlo, was developed in collaboration with researchers at Philips and Princeton University and predicts how cells send signals across networks to cause cancer or other disease. Details about the new method were recently published in Oncogene.

“Cellular signaling networks are the mechanisms that cells use to transfer, process, and respond to biological information derived from their immediate surroundings,” said Vinay Varadan, PhD, assistant professor at Case Western Reserve University School of Medicine, member of the Case Comprehensive Cancer Center, and senior corresponding author on the study. “InFlo can be viewed as modeling the flow of information within these signaling networks.”

InFlo works by assessing gene activity levels in tissue samples and predicting corresponding protein levels. It then uses statistical probabilities and other mathematical models to build activity webs showing how the proteins interact. Researchers can use InFlo to compare diseased and healthy tissues and pinpoint signaling differences. InFlo is tissue-specific and accounts for genetic alterations associated with disease, unlike other methods. It represents a major step forward in deciphering the activities of multi-tiered signaling networks commonly used by cells.

“Complex diseases such as cancer involve the simultaneous disruptions of multiple cellular processes acting in tandem,” said Varadan. “We developed InFlo to robustly integrate multiple molecular data streams and develop an integrative molecular portrait of an individual cancer sample.”

InFlo incorporates data related to each level of cell communication within a single sample, including DNA, RNA, proteins, and molecules commonly attached to proteins such as chemical methyl groups. The method also includes strategies to reduce “noise” and only highlight the signaling networks most likely to cause disease.

Analisa DiFeo, PhD, senior co-corresponding author on the study, Norma C. and Al I. Geller Designated Professor of Ovarian Cancer Research at Case Western Reserve University School of Medicine, and member of the Case Comprehensive Cancer Center, validated InFlo using ovarian cancer tumor cells that were resistant to platinum-based chemotherapy. InFlo pinpointed the interaction between two proteins called cAMP and CREB1 as a key mechanism associated with platinum resistance.

“Following up on InFlo’s predictions, we showed that inhibiting CREB1 potently sensitizes ovarian cancer cells to platinum therapy and is also effective in killing ovarian cancer stem cells. We are therefore excited about this discovery and are currently evaluating whether this could lead to a potential therapeutic target for the treatment of platinum-resistant ovarian cancer,” said DiFeo.

InFlo is being incorporated into Philips IntelliSpace Genomics platform, and will soon be available for widespread use in basic and translational research settings. Case Western Reserve University researchers will continue to develop the IntelliSpace Genomics InFlo module and the next step will be to expand InFlo to incorporate other data streams. “We are currently collaborating with the Imaging Informatics research group in the Center for Computational Imaging and Personalized Diagnostics at Case Western Reserve University to integrate InFlo with imaging-features derived from pathology and radiology data,” said Varadan. Such an addition would result in one of the most comprehensive tools available to researchers to infer mechanisms underlying complex diseases such as cancer.

Citation: Dimitrova, N., Nagaraj, A.B., Razi, A., Singh, S., Kamalakaran, S., Banerjee, N., Joseph, P., Mankovich, A., Mittal, P., DiFeo, A. and Varadan, V., 2016. InFlo: a novel systems biology framework identifies cAMP-CREB1 axis as a key modulator of platinum resistance in ovarian cancer. Oncogene.
DOI: http://dx.doi.org/10.1038/onc.2016.398
Research funding: Career Development Program of Case GI SPORE, Career Development Program in Computational Genomic Epidemiology of Cancer, Philips Healthcare, Rosalie and Morton Cohen Family Memorial Genomics Fund of University Hospitals, and others
Adapted from press release by  Case Western Reserve University School of Medicine

fABMACS – a new computational tool to simulate drug protein interaction

A new computational tool called fABMACS is helping scientists see beyond static images of proteins to more efficiently understand how these molecules function, which could ultimately speed up the drug discovery process.

Simulation of BRD4 interacting with a potent inhibitor.
Credit: Dr. Bradley Dickson, Rothbart Laboratory
Van Andel Research Institute

“The goal of targeted drug design is to create a molecule that interacts specifically with a protein, and this requires a description of protein-drug interactions that is precise–down to the placement of each atom,” said Bradley Dickson, Ph.D., a computational biophysicist in the laboratory of Scott Rothbart, Ph.D., at Van Andel Research Institute (VARI) and first author of a paper describing the tool. “The creation of fABMACS is a significant step toward robust virtual drug discovery because it saves time and money. It allows us to better harness the power of existing software while greatly improving our ability to predict the way that a potential drug interacts with a protein.”

Scientists often rely on collecting snapshots of proteins to determine how they may interact with a potential drug. However, these images are static and do not depict changes in proteins’ shape. “These snapshots provide valuable insight that can be enriched by fABMACS,” said Rothbart, assistant professor at VARI and the study’s senior author. “fABMACS allows us to simulate chemical changes to the drug and more quickly predict how those changes impact its interaction with the target protein. Ultimately, this could translate to improved drug potency and efficacy.”

To demonstrate the tool’s capabilities, the team ran several accelerated computer simulations of the epigenetic regulatory protein BRD4 bound to a drug that is currently in phase I clinical trials for blood cancers. They demonstrated that a slight change to the compound’s chemical structure could improve binding to its target protein, thereby improving its effect. The results of this work were published recently in the Journal of Chemical Physics.

fABMACS is an add-on to existing molecular dynamics software. It is based on GROMACSv5.0.5 and optimizes network communication and load balancing–both critical aspects of software development in parallel computing environments–to achieve a low-overhead implementation of new free-energy techniques. fABMACS also comes with a built-in configuration tool that allows the code to be tailored to different applications without requiring the user to manually edit the code, which maximizes transferability.

Citation: Dickson, Bradley M., Parker W. de Waal, Zachary H. Ramjan, H. Eric Xu, and Scott B. Rothbart. “A fast, open source implementation of adaptive biasing potentials uncovers a ligand design strategy for the chromatin regulator BRD4.” The Journal of Chemical Physics 145, no. 15 (2016): 154113.
DOI: http://dx.doi.org/10.1063/1.4964776
Adapted from press release by Van Andel Institute