Application of Complexity Theory: Away from Reductionist Phase Transitions

Reductionism persists as a useful node in the possibility space of understanding and managing the world around us. However the possibility space is now expanding to higher levels of resolution such as a focus on complex systems. Learning and tools are ratcheting in lock-step.

Some of the key complexity-related concepts in understanding collective behavior in real-life physical systems like the burning of a forest fire include:

• Organization and Self-Organization: Self-orchestration into order in both living and non-living systems, for example: salt crystals, graphene, protein molecules, schools of fish, flocks of birds, bee hives, intelligence and the brain, social structures 
• Order and Stability of Systems: Measurements of order, stability, and dynamical break-down in systems such as entropy, symmetry (and symmetry-breaking), critical point, phase transition, boundaries, and fractals (101 primer)
• Tunable Parameters: An element or parameter which doesn’t control the system, but can be tuned to influence the performance of the system (for example, temperature is a tunable parameter in the complex system of water becoming ice) 
• Perturbation and Reset: How and how quickly systems reset after being perturbed is another interesting aspect of complex systems 

 

Complexity science is not new as a field. What is new is first, a more congruous conceptual application of complexity thought in the sense of appreciating overall continuum of systems phenomena, not trying to grasp for the specific moment of a phase transition. Exemplar of this more comprehensive systems level thinking is Marcelo Gleiser’s reframe of the Grand Unified Theory problem and Sara Walker’s reframe of the Origins of Life problem. The other aspect that is new is the idea of working in an applied manner with complex systems, particularly with tools that are straightforward to implement like the math tools of non-linear dynamics, networks, chaos, fractals, and power laws (many inspired by the work of Stan Strogatz), and Software Tools like NetLogo, a multi-agent programmable modeling environment and ChucK, a digital audio programming language.

 

Citizen science health tools

The number of citizen science health and biology projects has been growing in the last few years due to a confluence of factors. Some of these include the plummeting cost of DNA sequencing, the availability of bioinformatics and other web-based data interpretation tools, the possibility of ordering direct-to-consumer blood tests, and having community DIYbio labs for experimentation, education, and support. DIYgenomics has developed a number of boilerplate tools to help in the design and conduct of citizen science health projects:

Study design and organization

• Study design template

Legal/ethics

• Sample informed consent form

Finance

• Study budget template

Recruitment and marketing

• Study flyers: multiple studies, aging study, vitamin deficiency study
• Conference poster
• Participant recruiting plan (per Genomera)
  

Human morphology-changing technologies

To date, most technology has been human-created. It can be grouped into two categories, technologies that are not likely to have an immediate direct impact on human morphology, and those that might.

Technologies that would likely not change human morphology
There could be the rapid advent of significantly more dramatic technologies than have been experienced to date. While these new technologies could change some aspects of life, human biological drives could remain unchanged, and therefore the structure and dynamics of human societal organization, interaction, and goal pursuit could also remain unchanged. Some examples of these advances could include the realization of molecular nanotechnology, quantum computing, cold fusion, and immortality. Even with several of these revolutionary technologies implemented, the seemingly different world would not actually be structurally different if humanity is still ordered around the same familiar biologically-driven goals.

Technologies that might change human morphology
The other group of technologies is those which could possibly have a near-term impact on the structure and form of what it means to be human, for example, cognitive augmentation, genomic therapies, and synthetic biology. The area with the greatest possible change is improving human mental capability. There have been several significant advances in a variety of neurology-related fields in the last few years that if ultimately realized, could potentially alter human morphology. Even the resolution of all mental pathologies such as Parkinson’s disease, depression, stroke rehabilitation, and addiction would constitute morphological change at a basic level. Augmenting cognition and deliberately managing biophysical states would constitute morphological change at other levels.

Morphological reach of technology

Is technology simply the tools that humanity has created to further its will or a force that can change humanity? Whether arrowheads or supercomputers, humans have made technology to enable and reinforce human nature and evolutionary tendencies. However, it is possible that advanced technology could actually change humans, human nature, and biological drives, both unintentionally and by design.

Aging research: systems biology, genomics and new tools

Three important themes emerged from the Buck Institute’s Systems Biology Symposium of Aging held November 10-13, 2009. The themes were progress in the overall understanding of aging as a systems biology problem, the role of genomics in aging, and new tools development for aging research. Happily, some immediately applicable tidbits were discussed: the findings of the protective response of endurance exercise, and the use of resistance exercise as a countermeasure to sarcopenia. (Mark Tarnopolsky)

Theme 1: Aging is a systems biology problem
Inflammation
Increasingly, aging is being understood as a systems biology problem involving cascades of signals across multiple pathways, many of which break down with aging. In younger organisms, problems are managed automatically as they arise, but in older organisms, the resolution processes do not work as well. When cells become damaged as a consequence of aging, they can either self-destruct through apoptosis (regulated cell death) or become senescent (living on without dividing). Senescent cells persist in tissues, where they may secrete inflammatory proteins. Many major age-related diseases, including atherosclerosis, heart attack, stroke and metabolic syndrome, share an inflammatory pathogenesis. The build-up of senescent cells can lead to both degenerative disease (aging) and hyper-proliferative disease (cancer). There are some efforts underway to facilitate the removal of senescent cells, for example, using an MMP inhibitor to kill senescent cells.

Dynamic regulatory continua
It is being suggested that more and more aspects of living systems such as humans are dynamic regulatory continua, and that there may be optimum points on the continuum which become harder to maintain with aging. One example of a dynamic regulatory continuum is the interrelation of cholesterol, fats, and Alzheimer’s disease. Having lower levels of the 142 alpha-beta plaques is neuroprotective, for example, but higher levels become harmful. One technique for understanding dynamic regulatory continua is to look at explaining the events at one biological level in terms of the events at the levels above and below them. (John Tower)

Signaling pathways
There is more of an effort to examine whole processes such as pathway networks and the chain of events in DNA transcription and translation. Current knowledge of signaling pathways is fairly primitive. The role of mRNA translation is being investigated as it is known to be related to growth promoting activities like cancer. There is the general translation of RNA, but this can be further modulated by the cell. In addition, signaling pathways are not working alone, there are probably many pathways converging. For example, there is likely cross-talk between several important signaling pathways such as the insulin pathway, the TGF-beta pathway, the IGF-1 pathway, and the TOR pathway. (Heidi Tissenbaum) In another example of the systemic interactions of aging, amyloid-binding compounds were found to suppress protein aggregation models in concert with homeostatic function (i.e., autophagy, chaperones, etc.). (Gordon Lithgow)

Theme 2: The role of genomics in aging
As with many areas of biology and medicine, the role of genomics is becoming increasingly important in aging. While it is known that there is little variation (0.1%) among SNPs in human genomes, 12% of the genome may vary structurally (copy-number variations, deletions, inversions and insertions of genes). On the threshold of whole human genome sequencing, it is being realized that SNP data alone is insufficient for a genomic understanding of health; more levels of data and annotated data, potentially including RNA sequencing to see protein expression will be needed. (Mike Snyder)

Variation in genomes
Three areas of research were presented regarding genome variation and aging. First were the long-expected results of Boston University's genome-wide association study (GWAS) on centenarians. The study found 150 SNPs in the genetic signature of longevity, 33 of which meet genome wide significance and are replicated. The most important longevity genes, most already associated with aging pathways, were: IL7 (immune system), CDKN2B (tumor suppressor), and APOE, CTNNA3, TOMM40, SORCS1, and SORCS2 (Alzheimer’s disease). (Tom Perls)

Related results were confirmed by personal genomics company 23andme. A study of senior athletes found that this cohort exhibited lower risk than the database in general. Ten chronic disease conditions were reviewed including coronary artery disease, breast cancer, prostate cancer, heart attack, type 2 diabetes, high blood pressure, high cholesterol, and macular degeneration. (Joanna Mountain) However, other research found that there is not a full overlap between genes conferring longevity and genes conferring increased healthspan. (Monica Driscoll)

Variation in genomic expression
Four interesting research findings found variation in genomic expression between older and younger organisms. First, another centenarian study found significant diversity of microbial communities in different age groups. For example, there was a high level of expression of certain miRNAs in older livers (miRNA-200c, miRNA-141, and miRNA-31). (Claudio Franceschi) A second study found that a full third of genome expression changed with age in worms. (Simon Melov)

A third study found a general relaxation in translational control and protein production during aging. It was proposed that increased or sloppy protein expression might contribute to proteotoxicity. (Monica Driscoll) Applying a systems biology and network analysis approach, a fourth study looked at how the structure of biological networks declines with age. The AGEMAP (a gene expression database for aging in mice) was reviewed, finding 26% fewer edges (edge nodes on the network) in 24 month old mice vs. 16 month old mice. It is possible that gene expression networks could lose integrity with age. An unexplored but possible explanation is that if there if less transcription, then network edges disappear. (Daniel Promislow)

Theme 3: New tools development for aging research
New approaches and tools are critical to advancing the study and potential remedy of aging, and three interesting talks were presented. First, progress in microfluidics and microscopy was discussed, particularly an exceptional development in electron microscopy that may allow the noninvasive molecular-resolution imaging of live samples (Figure 1). Usually electron microscopy is a destructive technique as the electron beam destroys the sample in the process of inspecting it. (paper: Noninvasive Electron Microscopy with Interaction-free Quantum Measurements). (Fatih Yanik)

Figure 1: In vivo noninvasive molecular imaging.

Image credit: http://www.rle.mit.edu/bbng

A second area of improvement has been in the targeted analysis of specific proteins. Now that there are robust measures for mRNA, proteins and post-translational modifications are the next areas of interest. Traditional shotgun analysis techniques are being improved upon by targeted analyses of specific proteins using mass spectrometry. The process is to take a protein mixture, produce peptides through proteolysis, collect a snapshot of multiple peptides at once, and use mass spectrometry to separate them by their mass. This method greatly expands protein identification and analysis capabilities, including the ability to do time course experiments. (Mike MacCoss)

Third, a genomic database tool, PharmGKB, was presented. The database facilitates a systems approach to pharmacology. Researchers can search for pharmacogenes, for example, given a drug and putative indication, ranking all genes in the genome for the likelihood of interactions. The database contains information regarding over 500 drugs, 500 diseases, and 700 genes with genotyped variants as of November 2009. (Russ Altman)

Bio-design automation and synbio tools

The ability to write DNA could have an even greater impact than the ability to read it. Synthetic biologists are developing standardized methodologies and tools to engineer biology into new and improved forms, and presented their progress at the first-of-its-kind Bio-Design Automation workshop (agenda, proceedings) in San Francisco, CA on July 27, 2009, co-located with the computing industry’s annual Design Automation Conference. As with many areas of technological advancement, the requisite focus is on tools, tools, tools! (A PDF of this article is available here.)


Experimental evidence has helped to solidify the mindset that biology is an engineering substrate like any other and the work is now centered on creating standardized tools that are useful and reliable in an experimental setting. The metaphor is very much that of computing: just as most contemporary software developers work at high levels of abstraction and need not concern themselves with the 1s and 0s of machine language, in the future, synthetic biology programmers would not need to work directly with the Ac, Cs, Gs and Ts of DNA or understand the architecture of promoters, terminators, open reading frames and such. However, with synthetic biology being in its early stages, the groundwork to define and assemble these abstraction layers is currently at task.

Status of DNA synthesis
At present, the DNA synthesis process is relatively unautomated, unstandardized and expensive ($0.50-$1.00 per base pair (bp)); it would cost $1.5-3 billion to synthesize a full human genome. Synthesized DNA, which can be ordered from numerous contract labs such as DNA 2.0 in Menlo Park, CA and Tech Dragon in Hong Kong, has been following Moore’s Law (actually faster than Moore’s Law Carlson Curves doubling at 2x/yr vs. 1.5x/yr), but is still slow compared to what is needed. Right now short oligos, oligonucleotide sequences up to 200 bp, can be reliably synthesized but a low-cost repeatable basis for genes and genomes extending into the millions of bp is needed. Further, design capability lags synthesis capability, being about 400-800-fold less capable and allowing only 10,000-20,000 bp systems to be fully forward-engineered at present.

So far, practitioners have organized the design and construction of DNA into four hierarchical tiers: DNA, parts, devices and systems. The status is that the first two tiers, DNA and parts (simple modules such as toggle switches and oscillators), are starting to be consistently identified, characterized and produced. This is allowing more of an upstream focus on the next two tiers, complex devices and systems, and the methodologies that are needed to assemble components together into large-scale structures, for example those containing 10 million bp of DNA.

Standardizing the manipulation of biology
A variety of applied research techniques for standardizing, simulating, predicting, modulating and controlling biology with computational chemistry, quantitative modeling, languages and software tools are under development and were presented at the workshop.

Models and algorithms
In the models and algorithms session, there were some examples of the use of biochemical reactions for computation and optimization, performing arithmetic computation essentially the same way a digital computer would. Basic mathematical models such as the CME (Chemical Master Equation) and SSA (Stochastic Simulation Algorithm) were applied and extended to model, predict and optimize pathways and describe and design networks of reactions.

Experimental biology
The experimental biology session considered some potential applications of synthetic biology, first the automated design of synthetic ribosome binding sites to make protein production faster or slower (finding that the translation rate can be predicted if the Gibbs free energy (delta G) can be predicted). Second, an in-cell disease protection mechanism was presented where synthetic genetic controllers were used to prevent the lysis normally occurring in the lysis-lysogeny switch turned on in the disease process (lysogeny is the no-harm state and lysis is the death state).

Tools and parts
In the tools and parts session, several software-based frameworks and design tools were presented, many of which are listed in the software tools section below.

Languages and standardization
The languages and standardization session had discussions of language standardization projects such as the BioStream language, PoBol (Provisional BioBrick Language) and the BioBrick Open Language (BOL).

Software tools: a SynBio CrunchUp
Several rigorous computer-aided design and validation software tools and platforms are emerging for applied synthetic biology, many of which are freely available and open-source.

• Clotho: An interoperable design framework supporting symbol, data model and data structure standardization; a toolset designed in a platform-based paradigm to consolidate existing synthetic biology tools into one working, integrated toolbox
• SynBioSS - Synthetic Biology Software Suite: A computer-aided synthetic biology tool for the design of synthetic gene regulatory networks; computational synthetic biology
• RBS Calculator: A biological engineering tool that predicts the translation initiation rate of a protein in bacteria; it may be used in Reverse Engineering or Forward Engineering modes
• SeEd - Sequence Editor (work in progress): A tool for designing coding sequence alterations, a system conceptually built around constraints instead of sequences
• Cellucidate: A web-based workspace for investigating the causal and dynamic properties of biological systems; a framework for modeling modular DNA parts for the predictable design of synthetic systems
• iBioSim: A design automation software for analyzing biochemical reaction network models including genetic circuits, models representing metabolic networks, cell-signaling pathways, and other biological and chemical systems
• GenoCAD: An experimental tool for building and verifying complex genetic constructs derived from a library of standard genetic parts
• TinkerCell: A computer-aided design software for synthetic biology

Future of BioCAD
One of the most encouraging aspects in the current evolution of synthetic biology is the integrations the field is forging with other disciplines, particularly electronics design and manufacture, DNA nanotechnology and bioinformatics.

Scientists are meticulously applying engineering principles to synthetic biology and realize that novel innovations are also required since there are issues specific to engineering biological systems. Some of these technical issues include device characterization, impedance, matching, rules of composition, noise, cellular context, environmental conditions, rational design vs. directed evolution, persistence, mutations, crosstalk, cell death, chemical diffusion, motility and incomplete biological models.

As it happened in computing, and is happening now in biology, the broader benefit of humanity having the ability to develop and standardize abstraction layers in any field can be envisioned.

Clearly there will be ongoing efforts to more granularly manipulate and create all manner of biology and matter. Some of the subsequent areas where standards and abstraction hierarchies could be useful, though not immediate, are the next generations of computing and communications, molecular nanotechnology (atomically precise matter construction from the bottom up), climate, weather and atmosphere management, planet terraforming and space colony construction.

(Image credits: www.3dscience.com, www.biodesignautomation.org)

Roadmap for Synthetic Biology

The most pressing issue in Synthetic Biology is building the groundwork to eventually advance to large-scale commercialization. How can the field’s growth from fringe to core be accelerated? A strategic plan for the Synthetic Biology ecosystem addressing academic, commercial, geopolitical and policy issues would help.

Academically, how many new bioengineering departments per year could be added? Open source course materials are available. Undergraduate and graduate bioengineering program templates including financing guidance, an industry association, faculty databases and implementation mechanisms are needed. Current academic conferences and journals could be expanded to reflect industry growth. There could be regional hands-on workshops for different levels of trained professionals and interested high-school students, similar to Math Jamborees.

Regarding enabling tools, there is a need for research and development, access ease, standardization and scale-up. Existing tools such as the PartsRegistry, OpenWetWare and Gingko Bioworks need to be taken to the next level. Academic and corporate research programs and incubators could develop a strategic roadmap for tools. An IEEE committee could be devoted to Synthetic Biology.

Commercially, there could be specific programs to involve the financing community. Venture Capital-backed SynBio Incubators could be initiated with conferences, programs, technology transfer and onsite startup incubation. Non-academic conferences, marketing and outreach programs, contests, prizes and X Prize grand challenge competitions could reside at incubators.

Safety protocols for practitioners and public discourse is a critical area for the success of Synthetic Biology. Asilomar and the Geneva Conventions could be helpful analogs for policy development.

Future of physics

The future of physics and cosmology was discussed at length at a recent conference held by the foundational physics research institute FQXi, particularly considering what may be ultimately possible and impossible for physics.

Theoretical physics has been progressing in many areas but there is still a strong need for observational evidence and/or alternative theories to support or disprove the existing ones. Luckily, much anticipated experimental evidence may be available in the next few years from the Large Hadron Collider (LHC), Planck Satellite, Pierre Auger Observatory and other observational astrophysics projects. For example, the Planck Satellite aims to look farther back in time than has been seen so far with the Cosmic Background Explorer (COBE), earlier than the 400,000 year old universe. It is theorized that B-modes, primordial gravitational waves from inflation, may be visible in the very early universe, which would provide additional proof of the inflationary phase occurring directly after the big bang.

Composition of dark energy to be known soon?
Another specific example for which observational evidence may be obtained in the next few years is regarding the composition of dark energy, whether it is vacuum energy made up of axion particles as one multiverse theory predicts or quintessence made up of supersymmetric WIMPs.

Many outstanding physics questions
Some other key issues are how our universe was created in the first place (a quantum theory of creation), proof for multiverse theories and possibly detecting bubble universe collisions, more about post-big bang inflation, why there is so much more antimatter than matter (the antimatter problem), why the weak force is 10³² times stronger than gravity (vs. say a more acceptable 3-4 orders of magnitude; the hierarchy problem), the existence, size and parameters of any additional dimensions of space, not to mention the usual unification of general relativity and quantum mechanics in a quantum theory of gravity, and finally the vexing Boltzmann brain problem, that consciousness could potentially arise from nothing but quantum fluctuations.

Next-gen astrophysics tools critical
Much progress has been made with accelerators, space-based telescopes, terrestrial array telescopes, and adaptive optics but the next era of astrophysics tools could be even more revolutionary. Accelerators are an expensive $5 billion or more and take years if not decades for fundraising, permitting, building and rendering operational. One alternative could be different kinds of accelerators which are smaller, quicker and cheaper to build, notably plasma wakefield accelerators, laser accelerators, and benchtop accelerators. Another way for tools to evolve could be with computational astrophysics and simulations such as MICA's Newtonian N-body simulation (featured by UgoTrade). As computing power continues to grow, accelerator and telescope datasets could be the inputs to large-scale simulation, prediction and test. As nearly every other science has moved into informatics and rigorous math-based prediction, simulation and experimentation, so too could astrophysics, fostering much quicker cycles and a tighter linkage between theoretical and experimental physics.

The FQXi community
Curiosity-driven physics researchers, especially those investigating risky areas on the boundaries of institutional acceptance are encouraged to apply for FQXi grants, and those understanding the value of fundamental physics research funding are encouraged to participate as donors and in the FQXi website community.

NOTE: The author is an advisor to FQXi.

Examining tool complexity

Tools and the science findings they enable evolve in lock-step. Many tools have been quietly transforming into complex entities of their own over the last several years. Exemplar contemporary tools on the landscape include many forms of the microscope, mass spectrometer, chromatograph, flow cytometer, and telescope.

The complex tools of today involve a hardware component together with many layers of software
for operating, enumerating and analyzing. The analytics software layer has become critical as mathematical modeling, simulation, automation, statistical computation and informatics are expected features. For example, the new biology extends traditional enumeration and experimentation with the additional steps of mathematical modeling and software simulation, and building test biological machines in the lab.

The increasing complexity of tools means that it is not possible to just wait for hardware speedups anymore, software is the weakest link (open source collaboration helps but only modestly), mathematical advances have been figuring most prominently and the cultural divide between hard science professionals and computer science, mathematics and statistical experts inhibits progress.

Alt approaches to AGI

50+ year old attempts at creating AGI have not been successful. It is possible that AGI cannot be generated from current methods and technologies; the wrong tool is being used, sort of like trying to build a 747 with a toothbrush. Electromagnetism, silicon and Von Neumann architectures may not ever have the capacity to achieve AGI even allowing for continued increases in processing, storage and memory and architectural shifts such as parallelism.

Other substrates might work
Getting around the rigidity of Von Neumann, mathematical, logic-based, computational approaches, symbolic approaches and traditional computers, other computational substrates like quantum computing, DNA computing, etc. might work and also those that humans have not yet invented, discovered or exploited for this purpose like light, air, memes and information. There must be other substrates, and other viable approaches that are not constrained by mathematics and logic.

Information as a substrate
Narrowly, the only existing example of general intelligence is the human brain and the basic requirements of AGI are self-replication and self-improvement. Considering self-replication, there are many examples of more effective self-replication than humans, for example, memes, disease and microbes. Considering self-improvement, memes also self-improve more effectively than humans as they are refined through repetition, and have the unbounded ceiling for improvement of true AGI.

Taking advantage of the self-reproducing and self-improving properties and using memes and information as a novel computing substrate might be one way of extending AGI progress.

Information as a substrate could be developed symbiotically with a very broadly applicable new understanding of the laws of physics based on information and entropy as opposed to mass and energy.

Pace of Encephalization

Humans have so much in their brains and so little shared outside. This has to change for true intelligence and species advancement and there is a great opportunity in developing tools to facilitate this.

1) Communication and interaction are currently limited by the mindset of the individual

Human interaction is fragile and conditions have to arrive at some optimum before meaningful communication can take place. A physical presence is generally though increasingly less required. Certain mixes of other people must be there or not be there and trust which generally must be developed over time must be present.

Only the few people in close proximity circles to others [may] have some level of understanding of what is in human minds. Peers, those that share similar ideas and values, may have a deeper but still sparse knowledge.

2) Communication and interaction are currently limited by language as the dominant tool

Communication is necessarily governed by the narrowband of language. Language was certainly an amazing evolutionary advance when it arose but it is time for new communication tools. Language is essentially a few pithy comments trickling out of multi-dimensional plane of existence or thinking on a topic. Other options to language that would allow the permissioned knowing of value systems, beliefs and history would contribute to enriched communications. Mechanisms for sharing clusters of thought rather than individual ideas would also be a start.

Tools
Two types of additional tools are needed: more tools for creating and sharing content and better tools for making the content meaningful. Content has proliferated but step function increases are coming. Humans will be creating and sharing more and more personal content (ideas, creative endeavors, personal life details, how-tos, resources, etc.) on the Internet via blogs, video blogs, tumblelogs, lifeblogging, interactive lifecasting, twittering, FaceBook, LinkedIn and other new methods. Aggregation, summarization and abstraction tools (like meta tag clouds and jaiku-style diverse feed aggregation) will be increasingly important to mark content relevancy and make it findable and interactable.

Just like businesses are wikinomically learning that they should not have boundaries at the edge of their properties, employees and ideas, individuals will hopefully start to realize the great benefits of extending their personal content boundaries.

Tools are Imperative

A continuation on the theme of humans being better innovators...

Creating new tools is often a key step in stimulating innovation. To do the innovation one wants, one may need to develop better tools to help. These tools in turn can beget an entirely new tier of innovation in a virtuous upward progression. A metaphor could be applied in some cases that the tools ARE the innovation much like Marsh McLuhan's medium IS the message.

Freeman Dyson cites the important example of astronomers, who build their own tools, being farther up the development curve than biotechnologists, who rely on others to build the tools they use. There appears to be a fundamental linkage and relationship between the tools and the innovation.

To innovate, we can focus on improving and creating tools. We can think of the areas of our lives and the tools we currently use there and problems we'd like to solve and what tools or solutions, objects, processes, etc. would do this and make life easier and better. We can dare to dream, create and invent!