Friday, December 16, 2011

Perfect class project for teaching science

Imagine a class project where the students have to find a few failed experiments (or experiments with confusing results) and develop hypothesis on possible explanations. Of course, some of them might be human error, but that is the part of the challenge of the assignment -- to develop reasons why that is the likely explanation.

I don't think this assignment is possible at present. Why? Below are my reasons, although I am sure there are many more.

1) There are a few journals of negative results, but I think they probably capture only a very small fraction of the number of failed experiments [in biological sciences as a whole].

2) EVEN IF all the failed results were published, how would a student navigate through them. Usually, understanding the nature of a failure is far far more difficult that understanding a successful experiment, because the pieces of a successful experiment make a coherent story. For a failed experiment, an expert in the field is often required to hypothesize potential explanations.

3) Failures MAY very often relate to each other. I, personally, think that most failed experiments can be attributed to a combination of human error and actual biological issues. If this is true, then looking at any single failed experiment is meaningless (sort of obvious). Finding "patterns" amongst thousands of failed experiments is practically impossible via the traditional journal-reading process.

Possible solutions (ambitious):

I believe that the human eye is extremely efficient in finding patterns, and therefore, it is necessary to leverage the visual + intuition capabilities of human researchers in order to make sense out of failed results.

A class project means that this task needs to be accomplished by non-experts. I think a framework that combines education + research + visual representation of results can by-pass this problem. In other words, provide ways to educate a student about a method when presenting the results from that method. If such a framework can be created, the boundary between education and research may become vague, which might be quite remarkable.


The traditional test-based system for selecting students is bound to miss several great minds who have the potential to solve many unresolved mysteries in science. If an infrastructure can be constructed where anyone can educate themselves and connect research results efficiently, there is a possibility that various geniuses around the world will be able to utilize their potential and service the community. The opportunity to circumvent the traditional test and money-based avenue to higher-education should provide an incentive for students around the world to try to do science on the side.

Thursday, December 8, 2011

Leave "pheromone trails" on research articles

What if there was a way to leave a "trail" (like ants) whenever researchers jump from one research article to another and find the connection very interesting. As more researchers walk the same path, the "pheromone" along that trail becomes stronger. This would provide a nice way of linking interesting research articles together, highlighting interesting patterns in the network of related research papers.

Wednesday, December 7, 2011

A dynamical systems analogy of society

Lets consider individuals in a human society as monomers. These monomers can interact in different ways to form larger complexes, with each complex having different functional capabilities. The reason why monomers might come together may be different. For example, in a modern company of workers, the monomers come together because of money - the salary provides the attractive force that brings the monomers together to form a larger cluster. Families are formed due to a different form of attractive force. Friends circle and so on have a different (maybe similar) force. Non-profit organizations or other occupations that are not motivated by salary (i.e. intrinsic reward based systems) attract monomers for an entirely different reason. I think this type of attractive force forms a larger variety of complexes, many of which interact more closely with the world outside human society (based on a few observations).

Now, lets consider society as a system with multiple stable states. Each stable state is defined by the types of complexes that exist in that state. For example, in one stable state, all the complexes are formed by salary-driven forces. In another stable state, the salary-driven forces are weak and therefore, intrinsic-reward forces are responsible for forming complexes. In another stable state, perhaps other forces are responsible for the complex formation. The question is: what causes these systems, i.e. societies, to shift from one state to another?

Tuesday, September 13, 2011

Yet another perspective on evolution

Consider this possibly true scenario:
Virus X lives inside 10% of the human population. It has integrated itself into the human genome, as many viruses do. It gets activated during the lifetime of a human and it exists the human body through the mouth and nose It is only able to invade other humans with specific proteins, which happens to be the same 10% of the population. In addition, this virus is also able to invade 10% of dogs and cats.
In the above situation, the virus X serves as a channel for DNA exchange. In other words, it is possible to build a "family tree" based on this virus. It is another a family tree based on a different means of reproduction. In normal reproduction, we do not reproduce our entire selves. Rather, some part of us is reproduced. Virus X is also reproducing a small part of us, and therefore, it is just like normal reproduction.

We think of the "tree of life" as a tree with lots of horizontal transfers. What if all these horizontal transfers actually form another tree of life, a tree that uses a different means of reproduction. Why limit to viruses? Microbes in our body are part of us. If microbes are mutating and being carried between species, then they, too, can be a means of reproducing a part of us. In summary, there might be several means by which we reproduce specific aspects of "us". Each of these means of reproduction can be used as a basis for building a different "tree of life". All these trees intersect.

So what is a species? It is basically a way of categorizing individuals. Just like there are different ways to classify books (e.g. language, subject, readers), there are different ways of classifying individuals. Each individual belongs in multiple species -- one for each type of reproduction process that affects the individual. We are all part of more than one species. So why do we give the regular form of reproduction so much importance? It's because of the brain, of course. The brain is not the master of our body. It is one component of the body, and just like any other component, it has roles. The brain is designed to pay attention to certain types of reproduction. For humans, these types of reproduction include sexual reproduction and cultural (meme) reproduction and maybe some other ones. We should not let the role of the brain affect the reality, which is that all of these forms of reproduction contribute significantly to natural evolution.

Therefore, natural evolution is a process where multiple trees of life are expanding and interacting. Individuals sit at the intersection points of these trees.

Wednesday, August 31, 2011

Crowd science

First of all, the two problems being addressed...

1) Great ideas are created when several small ideas come together. The
inability to keep up with a wide variety of research will lower the
chances of small ideas coming together. We're having trouble keeping
up with publications in just our own field, so imagine what all useful
discoveries we might be missing from other fields, especially those
obscure journals which might be hiding fragments of a great discovery.

2) Currently, if a curious high school student wants to learn more
about ongoing research, they would have to start reading review papers
or something like that -- this is not very inviting. No wonder the gap
between researchers and public is growing. The current mode of doing research requires a person to go through the PhD 4-6 year ritual. A self-motivated individual does not have sufficient material to become a self-motivated scientist. Even worse, a scientist who wants to cross from one field to another also needs a cross several large barriers.

And now, the solution...

An online game. Imagine a role-playing game where you walk around on
the earth (google earth-like interface with a person walking on it).
As your character walks through regions of the globe, you will see
little signs pop up indicating where the research labs are. There will
also be signs for "schools" (described later). When you click on a
research lab, you see comic strips of up-to-date research -- that's
right, comic strip of what each post-doc, graduate student, and
faculty did that day or any other day. You will also see the data
(graphs, tables, etc.) below each block of the comic strip.

Each comic strip is one line. It starts with a 'mission statement'
(objective or hypothesis) and ends with a concluding statement.
Conclusions can include emoticons, of course. Everything between those
two sentences describes what was was done using pictures and short
descriptions. The comic strips would describe procedures like miniprep
and PCR or even computational steps like parameter fitting. The player
can hover over the comic strip and find "schools" that teach those
concepts. E.g. I would hover over "PCR" step and see several schools
located across the globe that teach what PCR is, with ratings for each
school. Clicking on the schools takes me to online lessons (videos,
etc.) that teach those concepts. Schools with high ratings might even
make money from ads (incentive).

Ok, so now you ask how will these comic strips be generated... with
the Comic Maker of course! Each research lab participating in this
game can download a software called Comic Maker. Comic Maker comes
with hundreds of comic blocks representing some basic procedures. Each comic block can be generates by combining those basic blocks. It will have an easy drag-n-drop interface for creating a pipeline and attaching data to each step of the pipeline. The researcher must start a pipeline using a 'mission statement' and end it using some conclusion, even something as simple as :-( or just a few key words.

More fun stuff: researchers can announce "quests", which are open
problems that they are unable to resolve. Gamers can get involved in
quests. These players have to gather facts from other labs across the
world and generate some solution. They can request the researcher to
perform new experiments for them if they need more data. Similarly,
gamers can create novel hypotheses by collecting results from several
pipelines and present them to researchers.

One key question is: how does this reward the scientists who are in universities? Mainly, the reward is visibility, which is in many cases a big reward. Making research visible is a key for doing good research, and most researchers understand this obvious fact. Of course, placing results in this online game might interfere with the current "publications" approach. The workaround for that problem is simple: just place this in the game after they are published.

That's it. Hopefully that was a fun read. I think it can can be done.
Imagine spending your weekend looking at comics of what everyone is
going at Berkeley instead of going through the procedure section of a
paper. Of course, there will be nice search features, like "find me
everyone who is doing XYZ", where "XYZ" is some sequence of

Friday, August 19, 2011

Information vs information carrier

Suppose I write the word "Mango" using pen and paper. Then, suppose I wrote the same word using a chalk and blackboard. It is obvious that I am conveying the same information. The instrument used to convey that information is hardly relevant.

If this analogy can be applied to biological systems, then it is misleading to study physical aspects of signaling separately. Whether the signaling is via transcription factors, enzymes, RNA molecules, small metabolites, or DNA structure, there may not be any relevance with the content of the information that is being delivered by those molecular interactions. Similarly, between-cell communication may not be related to the type of signaling (paracrine, endocrine, quorum sensing, etc.). It might even be possible that the information encoded by molecular interactions is no different that information encoded by cellular interactions.

In summary, it is possible that information in biological systems might come to light if we study the patterns and ignore the physical components that create those patterns. At the same time, the physical aspects are not completely irrelevant. One would not write a book using chalk and blackboard, so the physical aspects of the instruments do restrict the type of information that can be conveyed.

Wednesday, August 10, 2011

Cell density based effects

In videos such as the one below, the cells in the middle have smaller size and therefore, there are more cells per unit area. This means that the concentration of molecules inside those cells might be different from the concentration of molecules in the cells at the outer boundaries. Suppose the cells have multiple stable states; in such cases, concentration differences can trigger state changes, causing different behavior of the cells based on their location in the colony.

Colonies where density is not equally distributed might use this fact (density dependent state change) to create different roles within the colony. Maybe the state of the cells at the center governs certain aspects of the colony and the state of the cells at the edges governs some other aspect of the colony. Even if evolution has not used this observation as a design strategy, it does not limit us from using the density difference as a design strategy for creating diversity within a population.

Sunday, June 26, 2011

Unnoticed evolution of humans

1) Suppose that the microbial community living inside us have a sufficiently long relationship with our own cells such that the microbes have an important role in a large majority of physiological functions, not just digestion.

2) As human societies have evolved, the variety of microbes that live in our environment have undoubtedly changed. Once humans wandered grasslands on bare-foot and ate raw food without even washing them. Now we buy processed food from supermarkets and even walk inside our houses with socks. The result is that the community of microbes that interact with and enter/exit our bodies has changed.

Combining (1) and (2) above leads to the possibility that the human physiology has indeed changed of the past several centuries, even though the change the human phenotype "appears" unchanged. The extent of evolution depends on the extent to which the microbial communities influence human physiology. If there are long-distance signaling molecules that are released by microbes which enter the blood and if the microbes are capable of receiving our signals, then...

Friday, April 22, 2011

Analogy for Protein Bursts

Consider this scenario:
Imagine a store that is open for a certain time interval during the day. In that time interval, several customers rush in to the store. Each customer buys several items. If someone would record the number of items sold as a function of time, he/she would probably observe bursts (one burst = items bought by one person). The time duration for which the store is open will correspond to the number of bursts. In other words, time duration maps to frequency of bursts.

Now, consider this transcription model:
Consider this mechanistic/intuitive model explaining how proteins are produced in bursts and how the frequency of bursts are controlled by the transcription factor:

1. transcription factor binds to promoter regions and opens the region for access by the polymerase
2. the region remains "open" for some time
3. during this time interval, the polymerase may initiate transcription multiple times
4. for each mRNA that the polymerase transcribes, multiple proteins are produced

So, in summary, the transcription factor "opens" the promoter region for the polymerase. Lets assume that upregulating the transcription factor affects the time duration of the "open" promoter. The longer the time interval, the more mRNA will be produced. Each mRNA creates a burst of proteins. Therefore, upregulating transcription factor affects frequency of bursts.

Stochasticity can lead to Stability

While we think of "noise" as a destabilizing force, we actually use noise many times to lead a system to the most stable position. Consider the scenario where you want to get all the items out of a hand bag very quickly; you would hold the bag up-side-down and shake, i.e. add noise. If you just tilt the bag up-side-down, there is a risk that some of the items would get stuck in the bag. Shaking ensures that such temporary "traps" (in mathematical terms, local minima) are avoided. In a sense, this is like stochastic optimization. When all the items are on the floor, they will remain there even if you shake the floor, because that is that is the most stable position.

Perhaps stochasticity in natural systems is a means of "shaking", i.e. a means by which natural systems reach the most stable states and avoid local traps. The most stable state remains relatively stable even at the presence of noise. Consider the situation when all the items from the bag are on the floor -- even shaking the floor would not change the situation significantly. Perhaps this is the nature of the "most stable state" -- it has a wide basin and thus can easily tolerate noise.

Friday, January 21, 2011

Centralized synthetic biology for the community

Build a "fun" framework that would attract lots of young minds to ask interesting questions, engineer biological cells to answer those questions, and easily share their results and explanations with each other.

1. providing the education to the masses... but I have some hope that young people are quite good are educating themselves, given the motivation
2. Resource
3. Safety

Possible solution:
One or two well maintained centralized robots with a fun interface for entering experiments and storing/visualizing the results. This is very much possible, except that it is expensive. But if the benefit is large enough, it is possible that government and/or companies might solve that issue.

But if such a programmable robot existed as a resource, I can imagine numerous researchers wanting to try new experiments and test new hypotheses. Since the data is centralized, there is a built-in benefit of data-integration, i.e. finding relationships between multiple experiments and drawing conclusions based on large number of experiments. It would be like a Facebook for biological experiments: everyone wants to dump their favorite hypothesis on it.

Thursday, January 20, 2011

Exercise as a general solution to living systems

I heard the statement a couple of times that places in this world with more cleanliness tend to have more auto-immune illnesses such as allergies. One can imagine that the immune system also needs to "exercise" from time to time, just like the muscles in our body or the brain.

What if "exercise" is a more general theme? Perhaps every pathway in a living system needs some form of exercise (but not too much, of course). What if the error-checking mechanisms and all the other anti-cancer mechanisms in our body also need exercise... perhaps we can cure cancer by exercising anti-cancer pathways. Exercise against cancer would be very tricky. It would require producing small controllable (or self-limiting) tumors in our body from time to time. By doing so, the hypothesis is that the body's anti-cancer pathways would remain alert (well exercised). This might make the body better prepared against the real cancer or tumor.

The same theme of exercise can perhaps be applied to achieve other goals, e.g. directed evolution of microbes.