Thomas Hunt Morgan, years before his famous flatworm research was published, reading papers at Columbia University.

Replicating Life in Code

<p>
 A few years ago, Michael Levin faced a conundrum. He and his colleagues at the Tufts Center for Regenerative and Developmental Biology just outside Boston wanted to find a model that would explain why the flatworm—a model organism used throughout biology—looks the way it does. At a fundamental level, they wanted to be able to describe the cascade of events that leads to the growth of a head in one place and a tail in the other.
</p>
<p>
 In fact, it was almost the same problem that Thomas Hunt Morgan, a Nobel Prize-winning evolutionary biologist and one of the founders of the modern study of genetics, faced more than a century ago. Back then, he was busying himself making careful cuts into flatworms. He sliced them lengthwise. He dissected them in half. From each segment, the worm grew a complete body. Eventually, Morgan carved off 1/279 th of the worm, a bit selected from its mid-section, and demonstrated that it could regenerate an entirely new animal. He was trying, unsuccessfully, to understand why and how certain body parts developed where they did: Why did a head appear at one end of the worm after a cut?
</p>
<div class="related"></div>
<p>
 Over the next 100 years, using new tools and insights, researchers replicated Morgan’s efforts in increasing detail: Make a cut here, a fully formed trunk and head will reappear. Tinker with this particular chemical or knock down this gene, and create a worm with a tail at either end. There are more than a thousand such papers. Yet, still, nobody has been able to fully explain why a head forms where it does.
</p>
<p>
 Levin and his colleagues at the center, where he’s the director, have been tackling this problem for years. In that time, they have helped explain some basic questions about development and regeneration. By tweaking certain signals within flatworms, for example, he has been able to grow a worm with four heads or one with no head but a tail on either end. He and his team have even grown an eye on a tadpole’s belly. But those experiments didn’t help them understand how it all fit together.
</p>
<div class="pullquote">
<p>
 “Maybe this is beyond the ability of us to come up with off the top of our heads.”
</p>
</div>
<p>
 “We have a massive literature of results saying ‘I did this to the worm and this happened,’ ” Levin says. “And we’re increasingly drowning in ever higher resolution genetic data sets. And yet, since Morgan cut his first worm, we still don’t have any model that explains more than a couple of different cuts.” And so, “after 120 years of really smart people going at it, I started to wonder, maybe this is beyond the ability of us to come up with off the top of our heads, to create a model that fits all the data.”
</p>
<div class="ad"></div>
<p>
 Computers, however, now offer enough power that Levin thought it might be possible to create such a model for a flatworm, one that would detail,
 <i>
  in silico
 </i>
 , the cascade of events that lead to the growth of a head at one end and a tail in the other. Still, even with a supercomputer’s worth of computational power, it wouldn’t be easy.
</p>
<p>
 Yet Levin’s ambitions didn’t stop there. Full, in-depth models beguile nearly every aspect of biology, and Levin hoped to create a tool that could be employed beyond flatworm development, one that could eventually model the vastly more complex world of human diseases. If he did it right, it might even help develop cures for those diseases. All he needed first was the right computer program.
</p>
<h2>
 Enter Evolution
</h2>
<p>
 It’s fitting that Levin’s computer models of biology are inspired by one of the fundamental tenets of biology itself, evolution. He builds his models around the idea that computer algorithms can meet, mate, and select the most fit version to then mate with other fit algorithms. These so-called genetic algorithms were first proposed by computer scientist John Holland in the 1960s. But even earlier, in the 1940s and ’50s, “people were already thinking about using inspiration from biology to build life-like computer programs,” says Melanie Mitchell, a professor of computer science at Portland State University and author of the book
 <i>
  An Introduction to Genetic Algorithms
 </i>
 . For instance, John Von Neumann, one of the earliest computer scientists in the 1940s, envisioned computers that could replicate themselves, with code serving as what we now call DNA. Mitchell says that Holland saw the field of genetic algorithms as a mathematical tool that could help explain how adaptation occurs in evolution.
</p>
<div class="pullquote">
<p>
 “Genetic algorithms can be used anywhere, from finance to rainfall in the Amazon.”
</p>
</div>
<p>
 What Holland saw as theory, others took as a practical tool. For instance, one of his students, David Goldberg, used these new genetic algorithms to optimize plans for gas pipelines by mating different models until the algorithm came up with the best design. But while early computer scientists were limited by memory and speed, today’s more powerful computers can run increasingly complex models to process millions of possible combinations and save the best chunks of code before passing them on to the next model, just as in natural evolution. Mitchell says these models have applications in engineering, big data, drug design, banking, and ever more realistic computer graphics and animations.
</p>
<p>
 Unlike biological evolution, where individuals meet, mate, and pass on useful traits that best fit the environment, computer-based evolution starts with a goal or a set of rules. Then the computer generates millions, even billions, of models to try to meet those goals. The ones that solve part of the problem or meet some aspect of the goal have a higher likelihood of passing on that relevant code to the next generation.
</p>
<p>
 Robotics researchers have employed this approach as well. Josh Bonguard at the University of Vermont modeled robots that learned to evolve walking. Columbia University’s Hod Lipson used this approach to simulate machines that learned how to crawl on a table. He spun off a company called Nutonian that allows scientists to input their data, and then the program evolves equations until one explains the data. That equation could allow researchers to optimize designs, model what might happen in the future, or show how changes in one part of a system might affect the final result. “It can be used anywhere,” Lipson says, “from finance to rainfall in the Amazon.”
</p>
<p>
 “This approach—reverse engineering—it’s like a Russian spy movie,” says Johannes Jaeger, a developmental geneticist and scientific director of the Konrad Lorenz Institute for Evolution and Cognition Research in Austria. “You have some kind of gadget, like in the Cold War, some kind of Russian technology. You don’t know what it does. Here, you have an organism, you don’t know how it works, and you’re trying to infer the networks that make the pattern that you see in animals.” Jaeger began working in this field more than a decade ago and has used such algorithms to model the genetic network that created a segmented body pattern in fruit flies.
</p>
<p>
 But none of the models are as complex as the development of an entire creature’s shape.
</p>
<h2>
 Billions of Experiments
</h2>
<p>
 When Levin first proposed his modeling project a few years ago, his colleagues in biology found the proposal absurd. “Pretty much nobody I talked to thought it was going to work,” Levin says. His critics had two overarching reactions. Some thought it would be impossible to find any model that worked; biologists would say, he said: “ ‘You’re telling me this program is going to take random models and by recombining random changes to random models you’re going to find the right model? That’s ridiculously impossible.’ ” Levin disregarded that criticism. That was how evolution had worked, he thought, and computers finally seemed powerful enough to try.
</p>
<p>
 The second criticism he heard was that they’d find many models that explained the data, maybe 10, maybe 1,000. How would they know which one was the correct one? “In theory,” Levin says, “that was a possible outcome. But we didn’t have
 <i>
  any
 </i>
 . It’s actually very difficult to find a model that does what it needs to. I wasn’t worried. If we found more than one—fabulous.”
</p>
<p>
 Levin hired post-doc Daniel Lobo to lead the flatworm modeling project, a computer science PhD who had worked with Hod Lipson, and whose research was also inspired by Johannes Jaeger. (Lobo now heads his own lab at the University of Maryland, Baltimore County.) Lobo had the right combination of computer expertise and interest in biology, and he’d written a paper about applying genetic algorithms to the evolution of shape that caught Levin’s attention. He’d used such algorithms to automatically design structures optimized for unmanned landings, such as those of the rovers sent to Mars, and Levin was impressed with the way Lobo combined a deep knowledge of the field with an interest in making it practical.
</p>
<p>
 The first challenge was to take the more than 1,000 experiments that had been done on flatworm shape and create one language to describe those results. It’s a not insignificant challenge. Natural language, as opposed to computer code, is ambiguous, even in scientific papers. At the same time, the team had to decide what to encode. They didn’t need exact dimensions of a flatworm head, but they did need to include relative scale, for instance. Eventually, Lobo created a standardized language and a standardized mathematical approach that represents the shapes of the worm’s regions, its organs, and how they’re interconnected. The end result was a searchable database of results, which is now available to all flatworm biologists.
</p>
<div class="newsletter"></div>
<p>
 Next, Lobo designed the simulation itself, a virtual worm on which candidate models would test their results. The computer compares the results of the simulated experiments to the real-world results expected from the database. The models receive scores based on how well they predict the outcomes seen in flesh-and-blood flatworms. Those with high scores reproduce; those with bad scores are discarded.
</p>
<p>
 After four years of work, they’d come up with a common language for scientific papers, distilled the most crucial aspects of the worm to put in a model, developed a simulator, and created the algorithm to find the shape model. Levin felt fairly sure that the project would work—but they needed one more piece of equipment. Common lab computers, no matter how powerful, can’t yet quickly process the massive computations needed to evolve an entire biological model, in effect replicating millions of years of evolution. They needed a supercomputer. So the team rented time on Stampede, the University of Texas supercomputer that can perform up to 10 quadrillion mathematical operations per second.
</p>
<p>
 Levin says the first models performed terribly; they didn’t get anything right. But by the 100 th generation, some of the models started to predict some of the correct results. By the 1,000 th , the models were increasingly matching the real-world experimental results.
</p>
<p>
 In the end, it took 6 billion simulated experiments, 26,727 generations of models, and about 42 hours of processing by the Stampede computer before the computer came up with one result. This was what they’d been waiting for: one model that could explain the 1,000 existing experiments that generate the head-trunk-tail pattern in a flatworm.
</p>
<p>
 To test the model, the team introduced data from two papers on shape formation that had purposefully not been included in the dataset when developing the models. The model accurately explained the results of both papers.
</p>
<p>
 So Levin and Lobo took it one step further. The model predicted experiments that had never been done before. The team tested those experiments in the real world, and they worked. The new results were published in May, 2016, describing the activity of a previously unknown gene that played a role in shape formation.
</p>
<p>
 Intriguingly, the model suggests the existence of a second node that hasn’t yet been explained by current scientific knowledge; it could be a protein, it could be a particular chemical. “The computer knows there’s a product that should be there, that seems to be important. In a way, it’s predicting a product that we don’t yet know,” Lobo says.
</p>
<h2>
 From Silicon Models to Hard Data
</h2>
<p>
 The shape investigation a success, Levin and Lobo turned their attention to modeling disease. They started with melanoma—skin cancer—and did so by focusing on pigmentation cells in tadpoles.
</p>
<p>
 They conducted experiments in which tadpoles were exposed to particular chemicals during their development. The tadpoles had no obvious chromosomal damage or genetic trigger, but for some of them, the chemicals would spark a change in all of their pigmentation cells. Those cells then turned metastatic and invaded tissues throughout their bodies. But for other tadpoles under exactly the same conditions, nothing changed. Could the algorithm determine why?
</p>
<p>
 Graduate student Maria Lobikin conducted dozens of experiments, knocking out genes or applying drugs and determining what percentage of the tadpoles became hyper-pigmented. Then she combined those results with other research published in the last decade. The team followed the same approach, creating a standardized language to describe the experiments and using a supercomputer to evolve a model to understand how, in some tadpoles, under certain biological circumstances, the cells flipped to a hyper-pigmented, cancerous state.
</p>
<p>
 The computer-generated model came quite close to the results of the existing experiments. The model predicted the results of all papers but one; perhaps the algorithm had even caught inaccurate data, a finding that was published in the journal
 <i>
  Science Signaling
 </i>
 in October 2015. “I said, you know what, go back and redo this experiment just to make sure, and sure enough, the data were slightly off,” Levin says. “It’s almost like a verification step. If it’s having trouble matching the results of one experiment, maybe the problem’s not the model, maybe the problem’s your data.”
</p>
<div class="pullquote">
<p>
 The simulated experiments demonstrate how artificial intelligence can augment human abilities.
</p>
</div>
<p>
 More recently, they asked the model a question. Is there any way to create a scenario where only some of the pigment cells become cancerous, where it’s not an all-or-none response? The computer generated an unusual three-step combination of drugs. When the team tried the experiment suggested by the computer model, they were able to create the first partially-pigmented animals.
</p>
<p>
 While the tadpole model is far from an ideal surrogate for human disease, Levin points out that this research supports what other scientists have claimed, that cancer is not always a result of specific DNA damage. Rather it may also be a systems disorder, where the exact right set of circumstances in the system generate conditions for cancer to grow.
</p>
<p>
 The simulated experiments demonstrate how artificial intelligence can augment human abilities, both Lobo and Levin say. “I see it definitely not as a way to replace biologists,” Jaeger says with a laugh. He says it’s nearly impossible for humans to process all the relevant parameters to generate a model, but computer power can do what our brains simply can’t.
</p>
<p>
 The duo see many more such experiments in the future. At his Baltimore lab, Lobo is now focusing on bacteria, as modeling the ways in which the microbes create different compounds could be useful for the field of synthetic biology. He’s also trying to reverse-engineer cancer tumors to attempt to discover the best possible treatments to cause them to collapse. Levin sees applications in many fields: drug development, regenerative medicine, and understanding metabolism and disease. (Levin recently was awarded one of the first two Paul Allen Frontiers Group grants, a $30 million grant over eight years, to support risky, unconventional research; these computer-generated models are only a portion of his lab’s research.)
</p>
<p>
 In any case, employing computer algorithms to wrestle with yet unanswered questions in biology will, researchers say, only become more mainstream. Lipson says this approach is crucial: “We’re in a stage where biology is producing lots and lots of data,” he says. “But magically it’s not going to make sense out of nowhere. You need these types of systems to make sense of the data we have.” In other words, systems that mimic evolution—and might help us evolve solutions as well.
</p>
<p>
 Image credits: Alicia DeWitt,
 <a href="https://hpsrepository.asu.edu/handle/10776/2122">
  Alfred F. Huettner/Marine Biological Laboratory Archives (CC BY-NC-SA)
 </a>
 ,
 <a href="https://commons.wikimedia.org/wiki/File:Smed.jpg">
  Alejandro Sánchez Alvarado/Wikimedia Commons (CC BY-SA)
 </a>
</p>

A graduate student tends cassava in an experimental plot.

The Next Green Revolution May Rely on Microbes

<p>To feed a planet of 9 billion, scientists are breeding mycorrhizal fungi that promise to boost crop yields by unlocking more nutrients in the soil.</p>

Networks of genes and proteins, like this network derived from human chromosome 22, could change the way we understand disease.

The Diseaseome Could Take Medicine Beyond the Genome

<div class="related"></div>
<p>
 When scientists first published the initial results of the human genome in 2001, we seemed to be on the precipice of a revolution in medicine. Researchers could finally discover specific genetic mutations that lead to diseases. Pharmaceutical companies could devise scores of new drugs to target those mutations. Patients could be treated based on their individual genetic patterns.
</p>
<p>
 So far, though, some scientists say the results have been disappointing. True, sequencing has become increasingly inexpensive, and this has led to a trove of data. Some of that data has provided important information to help prevent diseases; for instance, women who have the BRCA mutation and an increased likelihood of breast or ovarian cancer may be either carefully monitored or undergo a prophylactic mastectomy. Scientists have been able to use genetic sequencing to identify and group previously unknown diseases, which can lead to improvements in treatment.
</p>
<p>
 But the flood of new drugs based on the genome hasn’t arrived. “It’s been an unmitigated failure from my perspective,” says Joseph Loscalzo, head of the department of medicine at Brigham and Women’s Hospital and Harvard Medical School, referring to so-called genome-wide analysis studies that mine genetic data in the search for new targets for drugs.
</p>
<p>
 Loscalzo admits that the scientists involved may disagree with his forceful statement, but he points out that very few diseases have been linked to only one or a small number of genes. Even ones that are classically tied to a person’s genetics, such as sickle cell anemia, can cause vastly different symptoms, due, says Loscalzo, to the “genetic context” in which the mutated gene operates. For more complex and prevalent diseases, like cardiovascular disease—Loscalzo’s specialty—few patients have the same genetic variation or even the same symptoms.
</p>
<p>
 The problem, Loscalzo says, is that the approach is too simple. Our genomes are just one part of a vast network of interactions in our cells and in our bodies. The way to treat medicine in the future will rely on a new approach, he says, one that scientist are now beginning to flesh out: network medicine.
</p>
<h2>
 The Creation of the Network
</h2>
<p>
 László Barabási was visiting a Chicago playground in 2000 with his young son when he met fellow Hungarian Zoltan Oltvai. Barabási, then a professor of physics at Notre Dame, had a professional interest in network theory; he would soon co-author an influential paper and publish a best-selling book called
 <i>
  Linked: The New Science of Networks
 </i>
 . Olvai was a Northwestern University cancer researcher.
</p>
<p>
 Though he’d long been fascinated by biology, Barabási didn’t have the knowledge to delve deeply into the field. But on that Chicago playground, he and Oltvai began chatting. Oltvai suggested the two join forces.
</p>
<div class="pullquote">
<p>
 “A mechanic will never be able to fix the car without knowing how the parts interact with the other components.”
</p>
</div>
<p>
 Barabási was puzzled. “I said, ‘Well, where are the networks in biology?’ ” Barabási recalls himself thinking. But soon the pair published two important papers, one detailing the metabolic network in a cell, and another about the interactions among proteins.
</p>
<p>
 Barabási began to focus deeply on biological networks. He met Marc Vidal, director of the Center for Cancer Systems Biology at the Dana-Farber Cancer Institute, at a network science meeting, and the two hit it off. Barabási went on to spend a sabbatical year at Vidal’s lab in 2005. During that time, he connected with Loscalzo, with whom he now collaborates. He then started his own lab at Northeastern University’s Center for Complex Network Research.
</p>
<p>
 The genome hasn’t been a medical panacea, Barabási says, because it’s little more than a parts list, which can only get you so far. “A mechanic will never be able to fix the car without knowing how the parts interact with the other components,” he says.
</p>
<p>
 That’s the premise of network medicine. All the parts of a cell interact in certain way. Sections of DNA code for proteins, which can interact with other proteins, transcription factors can turn on stretches of DNA, and so on. It’s these connections and pathways—what some scientists are calling the “interactome”—that could help us understand how genes lead to disease. There are potentially millions, even billions, of such connections. But thanks to clever experiments and vast computing power, a map is starting to emerge.
</p>
<h2>
 Mapping the Links
</h2>
<p>
 Vidal has been focused on detailing the interactions among proteins for more than 20 years. Back in the 1990s, he was trying to understand how the genome as a whole codes for physical attributes such as eye color or asthma. He developed the idea that a crucial missing link is the network of connections between proteins in the cell, which interact to help translate the genome into physical attributes.
</p>
<p>
 At the time, he gave a presentation about this hypothesis. He imagined a map of such a wiring network, but, at the time, said it couldn’t be done. A colleague insisted that Vidal’s line of reasoning was too interesting to give up. “That night, I went to bed a little frustrated, thinking, ‘It’s a cool idea, but it’s impossible,’ ” Vidal says. “And the next day, I had an idea about how to solve it.”
</p>
<p>
 Vidal was an expert in yeast genetics, as yeast cells provide a simplified model that helps scientists tease out challenging questions about human cells. He employed a method to test many protein interactions simultaneously and relatively quickly to determine how they interlock. It works something like this: He begins by inserting yeast cells with sections of human DNA. Those sections then build the associated human proteins, as they would in a human cell. If the yeast cell the proteins created by those two of those DNA sections link up, then the proteins also interlock in our bodies.
</p>
<p>
 In 2005, Vidal published on the order of 1% of all possible human protein-protein interactions. This year, in a new paper currently under review, he and his team will publish approximately 20% of the expected full network. He’s since automated the system to speed the process and further improve the quality of the information. He hopes to produce a reference map of nearly the entire protein-protein network by 2020.
</p>
<p>
 Barabási, Vidal, and Loscalzo have been publishing papers together on such networks and nodes—the genes and pathways where diseases tend to cluster. They call such an approach the “diseasome.”
</p>
<p>
 Barabási likens the diseasome to a map of Manhattan: there are certain clusters where various activities take place—theater along Broadway, finance on Wall Street, advertising agencies on Fifth Avenue. The same is true for the patterns in a cell, though it doesn’t necessarily happen in physical space, the way neighborhoods cluster in a city. Rather, they play out in the chemical networks within a cell—specifically, which proteins, genes, and other chemicals are connected to one another. Using powerful computers to map these networks, researchers have been able to find regions where the connections that make up a particular disease “clump” together. They call these the “disease modules” in the network.
</p>
<p>
 Scientists have been able to find some of these modules. In a presentation at Dana-Farber Cancer Institute, Barabási described how the team of researchers had analyzed 300 diseases. They tried to determine whether the genes that are known to have a connection to a particular disease link up to each other through gene and protein interactions.
</p>
<div class="pullquote">
<p>
 Vidal hopes to produce a reference map of nearly the entire protein-protein network by 2020.
</p>
</div>
<p>
 They found that 20% of disease-related genes form a connected network in a disease module. The other 80% are in the vicinity; they connect to the diseased genes through one non-diseased gene. (This may sound familiar to fans of the pop-culture game “Six-Degrees of Kevin Bacon;” the 80% of genes in the vicinity are two degrees separated from the disease, just like Johnny Cash is two degrees separated from Kevin Bacon—Cash was in
 <i>
  All My Friends Are Cowboys
 </i>
 with Willford Brimley, who appeared with Kevin Bacon in
 <i>
  End of the Line
 </i>
 .) That means that if, hypothetically, there are 100 genes known to have a connection to asthma, about 20 of them link to each other. For the other 80%, Barabási believes that scientists probably haven’t yet discovered the missing connections that would link them up to the disease module.
</p>
<p>
 One surprising result that has come out of the networks thus far is an apparent relationship between asthma and celiac disease. Barabási and his colleagues demonstrated that there are similar pathways and significant overlaps in the wiring between the two diseases. Barabási says there’s also some clinical evidence that if you have one, you’re more likely to have the other.
</p>
<p>
 Aviv Regev, director of the Cell Circuits Program at the Broad Institute, has taken a different approach to understanding networks and diseases. Instead of building up known connections of proteins and genes that might occur in any cell, her team conducts a deep analysis of one type of cell. The group attempts to model just that one cell type’s wiring system, encompassing all the different chemicals and systems that link up, along with the non-coding elements of DNA. First, they focused on one variety of cell in our immune system that, when it malfunctions, is implicated in conditions including autoimmune diseases such as multiple sclerosis and inflammatory diseases such as Crohns and ulcerative colitis.
</p>
<p>
 It took years to understand all the connections in just that one cell. “But we developed the models from scratch,” Regev says. “Now the system is faster.” She adds, “We learned about unexpected players—including salt—that play key roles in autoimmune diseases.” They’re now taking the same approach with a wider variety of cells.
</p>
<p>
 The wiring that network medicine reveals will be crucial, Regev says. “Without figuring out that circuitry, it’s hard to figure out what genes actually do and act on that knowledge.” And though such data is far from complete, Barabási says, “If you have an incomplete map of Boston, it’s still very useful. It can still tell you how to get from here to there. There may be easier ways to get there, but it’s still predictive.” Even an incomplete map, he says, may help improve medical care.
</p>
<div class="ad"></div>
<h2>
 Finding the Network to Treat the Disease
</h2>
<p>
 Future medical treatments may not focus on a particular genetic mutation, but rather on the biological routes through which diseases are expressed. Such an approach could help in the case of certain melanomas, where patients with a particular genetic mutation were treated with a drug that inhibited that pathway. The melanomas receded, but within months, they returned full force. The cancer had found another pathway. By understanding melanoma’s network, doctors might employ many drugs that would act on many pathways, limiting its ability to evade treatment.
</p>
<p>
 James Collins, a professor of biomedical engineering at Boston University, agrees that the diseases of the future will be managed with drugs that act on multiple pathways in the network. (He admits that pharmaceutical companies, searching for one drug to treat a disease, haven’t been particularly happy to hear that; developing, say, four drugs for a disease instead of one could be significantly more difficult and expensive.)
</p>
<p>
 Another potential use of the diseasome could be in finding new uses for existing drugs. “Now we now what neighborhoods the diseases are in,” Barabási says, alluding to the genes and proteins that make up disease modules. “So, if a drug has been developed for heart disease but it tends to be hitting in the asthma neighborhood, it could be a good asthma drug.”
</p>
<p>
 Though Barabási predicts that network medicine will offer solutions in the near future, Loscalzo believes the approach is already yielding results. In fact, Loscalzo created the Harvard Medical School Division of Network Medicine in 2013 to investigate diseases revealed through network medicine.
</p>
<p>
 Working with the Barabási group, members of the Division of Network Medicine are trying to understand how changes in genes lead to disease—and to trace back through the network to find important genes that could be useful for treatment. They’ve been teasing apart the wiring of asthma. Loscalzo says they’ve used this approach to create subcategories of asthma. This could help determine, for instance, which patients might be more likely to benefit from specific treatments such as inhaled steroids, thus personalizing the asthma therapy.
</p>
<div class="pullquote">
<p>
 Using a network approach has already led to one success today: in the area of antibiotic-resistance.
</p>
</div>
<p>
 Loscalzo says they also identified a particular inflammatory pathway that led to pulmonary hypertension, where pressure in the arteries that carry blood from your heart to your lungs causes symptoms such as shortness of breath. The inflammatory pathway could serve as a potential drug target.
</p>
<p>
 Using a network approach has already led to one success in a particularly thorny area today: antibiotic-resistant bacteria. Collins’ team tried to work out the process by which antibiotics killed bacteria. Scientists thought they already knew the answer, but Collins says they didn’t have the entire picture. He and his team’s research demonstrated that the drugs trigger stress responses in cells. Due to that stress, the bacteria themselves eventually create damaging molecules, which harm their DNA and contribute to their death.
</p>
<p>
 But the chemicals that bacteria produce when they’re stressed do something else, too: they cause mutations in DNA that lead to antibiotic resistance. Today, antibiotic resistance is thought to emerge because, scientists have believed, there are a few bacteria in a given community that are naturally resistant to a drug, and they thrive after the drug kills off the bacteria’s brethren. But instead, as Collins’ research has demonstrated, antibiotics themselves induce mutations, leading to antibiotic-resistant bacteria.
</p>
<p>
 With this new understanding, Collins and his lab set out to combat antibiotic resistance. Collins’ lab introduced an extra protein to antibiotics that flips on DNA repair activity within the cells. Repairing the DNA prevents the mutations that create antibiotic-resistance in the first place. In studies, this process boosted the efficacy of a drug such as Cipro from ten times to a thousand fold. Collins has since founded a company, EnBiotix, to attempt to commercialize this approach and improve the efficacy of existing antibiotics. He expects that the newly improved antibiotics—based on a network approach—could be tested in clinical trials within the next couple of years.
</p>
<p>
 “Biology is complicated,” Collins says. The idea that scientists and drug companies can target all the physical expressions of a disease by going after just one gene is mostly wishful thinking, he notes. But he’s excited to be part of a growing community of scientists that are trying to understand all the pathways of a disease, among genes and proteins and the myriad other chemicals and signals in our bodies. And he’s “proud that network medicine is now leading to potential therapeutics.”
</p>
<p>
 Barabási likens the recent history of medicine to the time when maps of the world were rare. Back then, we had no idea how people and places could be connected. But, he adds, “once we had maps, it was completely transformative for business.”
</p>
<p>
 “In cell biology, there are no maps,” he says, “but they are starting to emerge.” And that could be completely transformative for medicine.
</p>
<div class="newsletter"></div>
<p>
 Image credit:
 <a href="http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0028213#pone-0028213-g010">
  Djebali et al. 2012,
  <i>
   PLoS ONE
  </i>
</a>
</p>

<p>The biochemical connections within our bodies may hint at treatments that could transform medicine.</p>

Cynthia Graber

Articles

Replicating Life in Code

The Next Green Revolution May Rely on Microbes

The Diseaseome Could Take Medicine Beyond the Genome

Support Provided By