An effort to extend the time between the recently learned and soon forgotten
Assorted lecture notes I scribbled into my lab notebook while listening to more knowledgeable people speak:
Gaddy Getz explained that of all the genes in the human genome ( there are roughly 20,500 ), somatic mutations related to cancers have been detected in 93%. The great majority of these mutations seem to be consequences of the cancers ("passenger" mutations), not causes. By comparison, only 254 genes have been identified as containing mutations that cause cancer ("driver" mutations). These numbers represent the most recent research, but are surely incomplete (many of the driver mutations were discovered only recently, and others will almost certainly be catalogued in the future). To examine what is currently known about cancer as it pertains to specific genes visit the following website.
Daniel MacArthur described a study which examined the DNA exome sequences of more than 200,000 individuals. The goal this time was to examine non-somatic (germline) differences between sequences and to draw associations between variants and disease. 91,000 of the sequences were processed through a particular pipeline, yielding 955 TB of data, all of which was eventually processed into a single file with a format specific to describing variants (a VCF file). In this file about 3 million mis-sense alleles were identified. 99% of these variants had a frequency of less than 1%, and 52% of the variants were seen only once. Since a mis-sense allele implies a modified protein, we can say that humans usually differ from each other in uncharacteristic ways (that is, not only is person A different from person B, but the ways in which person A differs person B are different from the ways in which person B differs from person C). While we each have our unique ways of being unique, however, there are also many commonalities: for example, the first three principal components of MacArthur's 91,000 sample data invariably represent ancestry. This fact doesn't make the stereotypical, media-fed notions of race into anything real, but it does mean that any two individuals sharing a common ancestry will differ from people of another ancestry in many ways that are characteristic. For more information on this study visit the following website.
We have two pellet stoves, and during the weekdays these stoves keep the house comfortably warm all winter long. When I'm home during the weekends, however, I prefer to build fires in the fireplace, creating a cozy, central locus around which people can gather. During a three-day weekend this winter I pulled out the bathroom scale to see if I could determine the amount of wood these fires consume. I stepped on the scale each I brought more wood to the fireplace, and the measurements lead to the following calculations.
In total I made 14 trips to the woodshed, hauling back an armload of split logs each time. These loads weighed an average of 15.8 kg, and altogether we burned 222 kg of wood over the long weekend. A few days afterwards I swept out the remaining ash, resulting in about 1.9 kg of fine gray powder. Combining these numbers I see that wood was converted into non-solid residual with an efficiency of about 99.1%, leaving behind only a small ash sack for the composting bin.
In terms of starting materials and final products the fire could be described this way:
Can these numbers tell me anything about the stove? Drawing from a few websites I Googled up (such as http://www.mha-net.org/, http://extension.oregonstate.edu/, and chimneysweeponline.com) it seems that the total heat available in typical household wood (containing 20% moisture) is about 15000 BTU/kg. If correct, then that number implies that my fancy, sealed fireplace with its pipes and blower is providing only about 10% efficiency. While 10% doesn't seem shocking (lots of heat must go up the chimney after all) it is also lower than I might've wished. Unfortunately I don't have a clear idea about whether a number such as 10% is typical for an operational, household stove working in non-theoretical conditions or whether I should try to find a way to do better.
How is wood heat from a cost perspective? In a typical year we've been burning 6600 kgs of cordwood (that's three cords at 2200 kg/cord), as well as 5442kgs of pellets (a truck delivers four 1.5 ton pallets each fall). The price for either kind of wood varies, based both on the quality of the wood and the supplier, but comes to something like $4000/yr to keep my family warm. Now I like the fires -- there's nothing better than sitting in front of one with my kids while it's cold and snowing outside. Furthermore it is somehow satisfying to know that every bit of warmth in the house each winter arrives via one or the other of my shoulders. That said, heating with wood doesn't seem to offer any clear economic advantage over oil or natural gas. That's unfortunate, since trees are fundamentally renewable while burning fossil fuels isn't. It would be better if our society provided financial incentives that favored sustainable methods of keeping folks warm in the winter.
Scientists at UC San Diego reported an interesting method last week which has some potential dangers. While the story is still unfolding, the broad outline is probably worth considering nnow.
Over the last two or three years a method referred to as CRISPR/Cas9 has exploded onto the scene. This technique allows researchers to cut and potentially rebuild segments of DNA with extraordinary specificity. Earlier techniques were largely hit-or-miss, and you might need to generate thousands of yeast colonies to get one (if any) that took up a DNA sequence of interest. The CRISPR system instead allows biologists to specify the sequence they want changed, and then to use some bacterial enzymes to change only that very small portion of the DNA. By itself this technique is an amazing breakthrough, opening up entirely new possibilities for the improvement of our understanding, and no doubt hastening the development of therapeutic agents and much else.
What happens, however, if you insert the CRISPR/Cas9 machinery itself into the genome? That is, don't simply modify the genome using CRISPR, but incorporate CRISPR itself into the genome, causing each piece of DNA to have the ability to modify other pieces of DNA? Since DNA is still confined to the cellular nucleus, so the CRISPR machinery can't go running through an organism changing every cell. It can, however, jump from one DNA strand to another, meaning if you start off aa heterozygote for this modification then you will become homozygous. This approach throws standard notions of Mendelian inheritance out the window and could very quickly fix itself in a population, and thereafter become impossible to remove from a population without killing every organism inheriting the mutation.
The idea is that this approach (deemed by its developers as a "mutagenic chain reaction") could be used to spread beneficial mutations. The authors of the study suggest that a beneficial mutation could thereby be spread, perhaps controlling some awful insect-borne pathogen such as malaria or Dengue fever. Surely ridding the world of Dengue fever would be an enormous step forward, and would reduce suffering and death in the tropical regions where such diseases are prevalent. This idea of spreading mutations outside of normal evolutionary mechanisms is also worrisome, however. If some other gene manages to become included in the CRISPR/Cas9 payload then we could see the rapid spread of other mutations in addition to the target mutation. I don't claim to know how to balance the benefits and dangers in this case, but I hope that everyone is proceeding with caution.
For the sake of reference, the news story which first brought this issue to my attention was reported here.