But this is the 1960′s. Computers are HUGE. Yes, yes, transistors had been invented years before–and are now in wide use. So, at least we’re not talking vacuum tubes. Egads. Tubes! Building computers means wiring a whole bunch of these transistors together. With wire. In other words, the world’s finest computers look a bit like that box of Christmas lights you don’t want to think about in the basement: tangled, ugly, mean and prone to failure if jostled. Not exactly conducive to placement in a rocket.

No biggie, you think. You’ll just have the computer on Earth–nice solid earth–radioing back and forth to the sensors and engines in the rocket. You can even correct for the speed-of-light delays! Problem solved! Light up some Lucky Strikes and call it a day.

But then you think of the return burn. If the Apollo craft are going to get out of lunar orbit and return to Earth, they’re going to need to fire the rocket engine on the far side of the Moon–you know, where radio waves can’t reach. Crap. I guess you’ll have to figure out a way of wiring together all those (4000 or so, egads!) transistors in a way that is small, light and durable enough to survive being rocketed into space. Time to create the first integrated circuit computer–father of every damn computer most of us have used, ever.

Better call MIT.

For the first time, scientists working on NASA’s Cassini mission have detected sodium salts in ice grains of Saturn’s outermost ring. Detecting salty ice indicates that Saturn’s moon Enceladus, which primarily replenishes the ring with material from discharging jets, could harbor a reservoir of liquid water — perhaps an ocean — beneath its surface.

Such an ocean would vastly increase the chance of life elsewhere in our solar system, beyond our own planet.

When I write about embryonic stem cells, I’m often saying something like this:

Making a new mammal requires a single cell to become hundreds of distinct cell types–each with a unique pattern of gene expression that is maintained throughout life.

The answer is tremendously complex–much deeper and interesting than “sperm meets egg.” Part of what’s going on in that first trimester is the establishment of all those hundreds of cell types. Complex three-dimensional geometry, a dozen or so of delicate signals, precise timing and luck itself all play into this process. It often fails–in a lab dish or in the gestation of a baby.

This figure lays out, to a rough approximation, our understanding of the various paths a differentiating cell can go down–step-wise–to reach all of those hundreds of cell types found in an adult. Start at the top, at a zygote (a fertilized egg). Trace your way through to all of the organ systems and tissues in the body. Embryonic stem cells, a kind of pluripotent stem cell derived from inner cell mass cells, act like they’re near the top of this tree; that is their utility.

The arrows are somewhat deceptive. Each of the branch points is a moment of chance. The developing fetus stacks the deck, but it’s still a game of cards. The arrows are more probabilities; they should look like Feynman diagrams, not such beautifully deterministic paths. (That’s the answer to the Jeopardy question ‘how can this figure be more complicated and daunting?’)

Any one of these steps can, and do, go wrong–and whole branches can be omitted by accident–rendering the development a failure.

The ideal response would be to replace the tissue and cells lost with new, full-functional replacements–regeneration. For the parts of our body that are constantly turning over–skin, blood, and to a lesser extent bone for examples–this is exactly what happens. Since the cells in these tissues are always being replaced anyways, injury is little more than a very bad day.

These tissues have resident adult stem cells. Stem cells, by definition, can divide and either make more copies of themselves or give rise to new functional cells. They do not do much by themselves. It makes inherent sense that a stem cell living in a given tissue can only make cells for that tissue; you don’t want bone to be replaced skin. Tissues that are undergoing constant wear and tear have stem cells.

Run down the leading causes of death in the United States and an interesting pattern emerges. Heart disease. Stroke. Diabetes. These are caused injuries to tissues that do not turn over as a regular matter of life. Heart, brain and insulin-producing cells in the pancreas lack functional adult stem cell populations to replace them when they’re lost. After injury, the body is stuck. It does what you’d do when faced with a broken car window and no replacement glass: duct tape. When no replacement functional cells are available, the body tends to repair itself by scarring over the area–replacing what used to be functional tissue with something tough and durable.

2500 Americans die each day from heart disease. The only way, right now, to replace the heart cells lost after a heart attack (or other injury) is whole-heart transplantation. Only about 2000 transplants can be performed each year–limited mostly by the availability of donor hearts.

Which brings us to the embryonic stem cell. We need replacement cells for these tissues. Embryonic stem cells, a kind of pluripotent stem cell, can become any cell type–including heart and brain cells. Only pluripotent stem cells, to date, have an unquestioned ability to become heart cells.

The generation of an embryonic stem cell line involves the destruction of an about 100-cell pre-implantation embryo. At this stage, the embryo looks like a beach ball filled with sand. The cells making up the hollow sphere are the trophoectoderm and can go on to become all the supporting structures of a pregnancy (placenta, amnion and such). The clump of cells at the bottom of the sphere is the inner cell mass. These are the cells that can go on to become any cell type in the body.

To generate a new embryonic stem cell line, the trophoectodermal cells are removed by immunosurgery, leaving only the inner cell mass cells behind. These cells are then placed in culture conditions that promote their division as pluripotent cells–retaining the ability to become any cell type in the body. From about fifty cells, trillions can be made. This is why most labs doing pluripotent stem cell research never work with embryos. Since the lines can divide (nearly) indefinitely, the overwhelming majority of research involves work with existing lines–not the generation of new ones.

No human embryo, to my knowledge, has been destroyed only to make an embryonic stem cell line. Every single line in existence was created from an excess IVF (in-vitro fertilization) embryo. IVF requires the mixing of human eggs and sperm in a dish. Sperm is easy enough to collect, store, freeze and thaw. Human eggs are another matter. The collection protocol is dangerous to the woman–involving massive dosages of hormones. Worse yet, there hasn’t been a reliable way to store, freeze or thaw unfertilized human eggs. Fortunately, pre-implantation human embryos can be frozen and stored. So when human eggs are collected, all must be fertilized more-or-less immediately. Some of the resultant embryos are implanted immediately, with the rest frozen for future attempts at having children.

When the couple has decided they’re done having children, the left-over excess embryos are generally destroyed. (Nobody wants to pay for their continued cryostorage.) Since IVF is unregulated, it’s hard to know how many frozen human embryos exist in the United States. Estimates hover around a half-million. A tiny number are donated to other couples seeking fertility treatment, with others donated to scientific research–to be used to create embryonic stem cell lines.

This is a key point: If these cells weren’t used to create an embryonic stem cell line, they would be destroyed anyways. And there are hundreds of thousands of embryos in storage today. Every month they spend in stasis lowers their viability.

Are these embryos, being destroyed as a consequence of in-vitro fertilization, human beings? This is, at its core, not a scientific question.

The embryos are undeniably made up of human cells. But, so are the many skin, gut, and other cells you shed or digest every day. Having the right number of human chromosomes seems a poor standard for a human being.

Nor are these embryos an individual yet. If you split a blastocyst in half, you can get identical twins. Mash two together and you can get a chimera. It seems that any standard for a human being would require the entity to be an individual. Blastocysts are not yet individuals.

These embryos, if implanted, have a potential to develop into a human being. But, it’s important to carefully consider how great this potential really is.

We can estimate there are about a half-million embryos stored in freezers around the country. Only 134 of these excess embryos have been ‘adopted’ by other couples. More are being generated every day. So, a given excess IVF embryo only has about a 1: 4,000 chance of being implanted and delivered to birth. About half of embryos attempted for IVF (again this is a very rough estimate) even manage to implant. Of those that implant, even more miscarry.

We finally come to the question of what makes a cell capable of becoming any other cell? The basic science I work on in the lab is somewhat related to this question; how does a single cell manage to become hundreds of distinct cell types–each with a unique pattern of gene expression that is maintained throughout life. It’s a huge question that I narrow by focusing on the path from anything to heart cell.

The answer is tremendously complex–much deeper and interesting than “sperm meets egg.” Part of what’s going on in that first trimester is the establishment of all those hundreds of cell types. Complex three-dimensional geometry, a dozen or so of delicate signals, precise timing and luck itself all play into this process. It often fails–in a lab dish or in the gestation of a baby.

For all these reasons, and more, I find it hard to accept that a human life starts before the end of the first trimester. Humans are not yeast or bacteria. It takes those three months to even get the vague shape of the complex machine of human life.

In fact, it’s not until deep into the third trimester when all of these cell types are formed into organs that function. Until late in gestation, independent human life is not possible; survival takes extraordinary mechanical life support.

Nor does reaching one step, in any way, mean the next is going to succeed. Very little about development is deterministic; much is subject to the whims of chance and environment.

Something else has emerged recently: induced pluripotent cells, or iPS cells. This technique–to turn some committed cells into a cell that resembles an embryonic stem cell–was developed in Japan by Dr. Yamanaka. Japan’s restrictions on embryonic stem cell research were (and are) far more restrictive than the US’s.

This reprogramming technique required, first, extensive study of human embryonic stem cells. Dr. Yamanaka used a list of potential master regulator genes that was generated based on a large amount of work done on embryonic stem cells.

And iPS cells aren’t anywhere close to perfect. My lab–and I specifically–have done experiments to compare these iPS cells to true embryonic stem cells. They aren’t the same–with a significantly hindered abilities. We’re attempting some further reprogramming techniques to make them better–but that’s still years off. For now, the best bet for making iPS cells work well enough to use for therapies is to continue studying them and embryonic stem cells–using what’s learned in the latter to improve the former.

When making an ethical judgment, I believe it’s critical for us to balance the interests of both the excess embryos produced by IVF (that I do not believe to be human beings) with those of the hundreds of millions of adult human beings that are currently suffering from horrible illnesses.

I do not believe human embryonic stem cell research should be a free-for-all. While I do not personally believe that human life starts at the moment when sperm-meets-egg, I do recognize that human blastocysts deserve serious treatment.

I believe that the donation of blastocysts and the distribution of the subsequent embryonic stem cell lines should be strictly decommercialized. The entire process should be like how we handle organ donation from adults–with oversight, and the prohibition of money changing hands in the process.

Obama’s easing of the Federal funding restrictions on human embryonic stem cell research opens the door for such a policy in a way that Bush’s restrictions never did. By preventing public funding, the destruction of human embryos was forced into the private sector. In a horrifying way, Bush’s policies made the destruction of human embryos a matter of private enterprise–a potentially for-profit venture. Anything else would be better.

Creationists–particularly those under the guise of intelligent design–love these. How could something as complex as the coagulation cascade simply appear, intact and whole, due to random chance? There are no intermediate steps where small refinements can be selected for! It’s all or nothing, baby! You know what that means….

Proof of GOD! Praise Jesus!

Really? Some new work, in the field of skeletal muscle differentiation, points to another way beyond the divine to generate a cascade.

If we’re going to figure out a way something could happen by evolution, it’s important to show each step has something can can be refined by selection for the most successful versions.

Let’s say the master regulator here (“Let’s make a birthday cake”) starts off as quite promiscuous, willing to turn on a whole set of tasks all at the same time–some things you’d want in cake, some things you wouldn’t.

The cake is going to be pretty bad. Things that do not belong in cake are going to end up in the finished product. Likewise, the order is all off. It would serve as a cake-like substance, but there’s much room for incremental improvement. For example, “apply ketchup” would be strongly selected against.

We also have something wrong here. The sub-regulators (“place candle” and others) also must be turning on a whole bunch of other smaller tasks. Why couldn’t they be turning each other on as well?

We get a cake again, and a little better better one at that. Now, over time these in-order connections between the sub-regulators will get selected for until they’re nearly as strong as the master regulator. (The out of order connections between the sub-tasks would be selected against; they’re in dashed green to indicate this.) If both the master regulator and the previous sub-regulator must be turned on in order to activate the next step, the result would be a pretty damn good cake. Therefore, natural selection will pick out those cascades with that trait.

The better ordered these steps, the better the cake turns out. So, the strength of the connections between steps, relative to that of the master regulator, will increase.

Eventually, the master regulator will only be able to turn on the first step. The result? A neat cascade:

This isn’t just speculation on my part. Scientists studying the differentiation of skeletal muscle cells have found direct evidence for this sort of evolutionary process. Pretty cool.

Early in 1986, I snuck out of bed and turned on the ancient black-and-white television set close to my bedroom. Only eight years old, I sat down to watch a PBS special. The Voyager 2 spacecraft was about to send back the first close-up pictures of the planet Uranus.

Even at that young age, I understood I was enjoying a legacy of a prior era, a gift from a different time for the country. This was Regan’s America, dominated by stupidity and greed, inward thinking and striving. It wasn’t a time suited for anyone bookish, nerdy and dreaming.

And yet, there I was lying on brown carpeting, watching excited engineers and scientists catch their first glimpse–anyone’s first glimpse–of a distant planet. Men and women has created this carefully machined work of functional art, and precisely calculated a trip skittering along the vast wells of gravity dotting our solar system. It was magnificent. I wanted to join those men and women, to bring other such creations into the universe.

Darwin’s major accomplishment was to condense a lot of thought on the origins of life into two basic concepts: new traits arise randomly (mutation) and the most adaptive of these new traits would become dominant in the population (natural selection)–forming the first cohesive theory of evolution.

For proof, in these early days, we had Darwin’s observations on the Galapagos Islands and the fossil records showing the rise of new traits in the living population to match changes in or introductions to new environments.

Building off of Darwin’s ideas of natural selection and mutation generating new traits came Mendel, and his conception of genetics, a systematic way by which traits are passed from parents to children. Mendel’s genes passed unchanged from parent to child cause traits of living things. An individual has two copies of each gene, one from each parent. If you have a mixture of genes for a trait, one of these genes can dominate over the other, hiding the weaker recessive gene’s trait for the generation.

Watson, Crick, Wilkins and Franklin‘s discovery of the structure of DNA in the 1950′s gave genes a physical manifestation—understandable with fairly simple chemistry. The central dogma of biology followed shortly after, in which DNA encoding for genes is transcribed into messenger RNA and in turn proteins that cause the traits first observed by Darwin hundreds of years before.

Human understanding of life has come in these spurts, separated by decades of consolidation and grappling with new data or new ways of thinking about biology. We’re, right now, in midst of another spurt in our understanding of life.

Until about a decade ago, we only really knew DNA–the long ordered strands of the four basic letters of life–in patches and spurts, with little sense of the overall map of any living thing. With enormous effort and cost, we sequenced the human genome–the vast majority of all the DNA in a human cell. We discovered that a human being has about twenty-thousand pairs of Mendel’s genes.

In the past few years, sequencing DNA has become shockingly less expensive. What once cost four billion dollars can now be done for a few thousand. And the price is dropping dramatically every few years. As a result, we now have sequenced the genomes of many other organisms–fish, cows, dogs, mice, opossums, frogs, chickens, chimps to name just a few.

If we think about Darwin’s traits as tasks a living thing must accomplish–eating, carrying oxygen and so on–and Mendel’s genes as means of accomplishing these tasks–a beak, a histone protein to wrap DNA around, hemoglobin in red blood cells–by comparing the DNA sequences for given traits (a gene’s locus) in different organisms, we can see how evolution has adapted each organism to its environment and lifestyle.

Storing DNA should be little different for a fish, frog, mouse or human. The task is as ancient as eukaryotic life. Let’s look at the genetic locus coding for a protein responsible for this DNA-storing trait, comparing how close various other organism’s DNA is to human’s sequence.

(click for a larger version)

Each row represents the equivalent gene in another species (zebrafish, xenopus frogs, chickens, opossums, mice, dogs and macaque respectively.) The higher the blue, the closer the DNA sequence matches that of humans; given that our common ancestor with these organisms were millions, hundreds of millions for the majority, of years ago, this is a very high degree of sequence conservation. The task hasn’t changed (the goals of the trait) so the gene hasn’t changed much either.

Now let’s look at the gene responsible for the trait of carrying oxygen in our blood, the beta-chain of hemoglobin. For a fish or a frog, the demands of the task of capturing and carrying oxygen are dramatically different than those for a purely land-dwelling animal; we’d expect the DNA sequence for the equivalent gene in these animals to be quite different from human’s.

Indeed, that’s the case:

Today, on Darwin’s 200th birthday, I encourage you to browse around, comparing human genes to those of our distant and close relatives in the animal kingdom. (Check out Opsin 1, for color vision, or the NMDA neurotransmitter receptor for some other fun genes.) You are living in a golden age of biology, in which our understanding of life is jumping by leaps and bounds every year. Within your lifetime already, we’ve gained astonishing abilities to peer into the nature, structure and function of life. Despite the centuries of advancement, Darwin’s (and Mendel’s) carefully crafted ideas of evolution and genetics have not only endured, but provided an invaluable map to understanding this vast new collection of data. Be proud.

FiveThirtyEight’s election-eve prediction, of 349 electoral votes for Obama:

Reality this afternoon, of a projected 349 electoral votes for Obama:

I might start caring about baseball, just to further appreciate the awesomeness of Nate Silver.

What’s the best way to display all this data? A histogram.
Here’s Nate’s:

Along the bottom, on the horizontal, are the possible electoral vote counts for Obama.

For each one, from zero to five hundred thirty eight, on the vertical are the number of times this Obama electoral vote count happened during his ten thousand simulations. The tallest peaks are the most likely outcomes during the simulation. The low tails are things that are possible, but not very likely.

Many of the closer followers of FiveThirtyEight.com, like the Stranger’s own Anthony Hecht, tend to focus more on the big Obama victory pie chart. Over the past few days, Obama’s number has drifted down a bit, from a peak around 96% to the low nineties today.

Look at the histogram for today:

McCain is all tail, no peak. The peaks are still strongly skewed to an Obama blowout.

The histogram tells you, in much more detail than an number or a pie chart, the chances of the different outcomes in crisp (and this case comforting) detail.

I love histograms.

If I were a woman accused of claiming my daughter’s child was my own, and I knew such accusations were false, I’d use science to prove myself right.

Using genetics to test for maternity is no different than in paternity testing, and I wrote about that a bit ago:

Paternity testing is all about mixing and matching.

DNA-based tests work by comparing the recipes—the alleles—you have for a given gene to those of a possible child. For most genes, we get two alleles—one from our mother and one from our father. A child of yours must have one of your alleles for all the paired genes in his or her DNA.

Well, how can we use this to figure out if a baby is from a woman or her daughter? Let’s look at the example for the pair of alleles for one gene:

(Click on the image for a larger version.)

In this example, I’ve assumed the most complicated combination–where everyone is heterozygous with distinct alleles. Since we know the relationship between the daughter and her parents, we must assume she received one allele from each parent. Therefore, we have six alleles, giving us 15 possible combinations. (Want to check my math? The formula is 6 choose 2.)

Try to follow the logic yourself. Of the fifteen possible babies, several definitively cannot be the mother’s child–as the baby doesn’t have an allele from the mother. Several are definitively not the daughter’s child (for the same reason.) A few combinations cannot be either of their child (mixup at the hospital)? Most are ambiguous–could be from either potential mother.

Well, how do we deal with the ambiguity? If you look at many genes at once (trivially easy to do with current technology), it become easy to rule out a potential mother.