The post Puzzled appeared first on The Perry Bible Fellowship.
A Massachusetts reader of the site sent along this paper, which is from a journal that I am sure I have never looked into (!) It's from a multicountry team looking at human connectivity and networks in several forms (social media friendships, physical moves between cities, and scientific co-authorships) and comparing those between the US and China.
For example, at right is the social media graph for China, displayed in two different ways: a geographic map and a minimum-spanning-tree network plot. The latter makes it extremely obvious that Beijing really does seem to be the Hauptstadt of China by this measure, since huge areas like Shanghai only show up as subsidiary notes compared to the huge centality of Beijing itself. If you go to the paper and look at the US version, New York is indeed in the center of the graph, but nowhere near as much as Beijing is in the center of the Chinese one. There are notable subsidiary nodes for Chicago, Los Angeles, Washington, Philadelphia, Atlanta, and several other cities, and there's more of a multibranched topology rather than the all-roads-lead-to-Beijing look of the Chinese graph.
I won't discuss the city mobility graphs, although they're certainly worth a look, because I wanted to highlight the scientific co-authorship ones. These were extracted from the Web of Science data as of 2017, and for the US, reported author zip codes were then matched up with cities and Metropolitan Statistical Areas (MSAs) to capture regions more usefully by not counting every nearby municipality as a separate city unto itself. At left you see first the minimum-spanning tree for China in scientific collaborations, and again, Beijing is at the center of it, and quite dominant. But you can see that the topology is different than the social media one, with larger cities/research centers connecting to Beijing and smaller ones being more compactly described as connecting to those.
Below that is the US scientific co-authorship graph, and will you look at Boston. At no other point in any of the graphs does a city that's a smaller player in the other plots suddenly leap into a position in the center like this. San Francisco makes a pretty good leap into prominence too, as you would expect, but Boston displaces New York and all the other larger population centers to take the center. As mentioned, "Boston" in this definition includes Cambridge, Watertown, Lincoln, Waltham, Lexington and all the other municipalities that carry university, technology, and biopharma research centers, but it has to be noted that "San Francisco" does not go so far as to include San Jose (which is a node off to its right). My impression is, though, that if you bunched the whole Bay Area together in this measurement that you still wouldn't displace Boston from the center. Washington surely benefits from the inclusion of Rockville, Bethesda, and so on as well.
This illustrates something that residents, businesses, and policymakers alike already feel that they know - that yeah, the greater Boston area is a pretty big deal in science and engineering - but I think it shows it in a more emphatic way than usual. It's a very unusual situation, and it would be wise not to take it for granted. As it is, people in all sorts of disciplines move to this area because of the employers and the opportunities, and the employers move here because of the opportunities and the people. It doesn't have to be that way, though. It would take some work to mess it up, since the momentum of Harvard and MIT and Brandeis and Tufts and Northeastern all the rest of them, as well as the industrial momentum of all the biopharma companies, etc., would take time to dissipate. As Adam Smith said after Britain lost the US colonies, "There is a great deal of ruin in a nation". On the optimistic side, which is where Smith was likely coming from by temperament (see The Theory of Moral Sentiments), you can read that as how difficult it is to mess something up that large and how such things are more resilient than you might imagine. But it can also mean that ruin is already in process, running slowly enough that you can pretend that all is well and that nothing needs to be addressed. It would be prudent not to lean too far into either of those viewpoints!
Rather, I try to remember and appreciate that I live and work in a very strange and unusual area, with a weird concentration of weird people. I like to think that I fit right in!
I wrote here about some of the controversy around the superconductor discovery claims from the lab of Ranga Dias at Rochester, and the story has not taken any positive turns since I did that post. It is widely believed that the various papers on this subject from that lab are not reproducible, and there have been several retractions so far. This piece at Nature will give you the details about that, and about how things got to this point.
How things got to this point was, it seems, a principal investigator steamrolling his own grad students and postdocs, hiding key details and pressuring everyone to get papers out no matter what misgivings they might be feeling.
By the time the CSH paper came under scrutiny by Nature journal editors in early 2022, Dias’s graduate students were starting to grow concerned. In summer 2021, Dias had tasked them with investigating a compound of lutetium and hydrogen (LuH), which he thought might be a high-temperature superconductor.
They began testing commercially purchased samples of LuH and, before long, a student measured the resistance dropping to zero at a temperature of around 300 K (27 °C). Dias concluded the material was a room-temperature superconductor, even though there was extremely little evidence, several students told Nature. “Ranga was convinced,” one student says. . .
But the measurements were plagued by systematic errors, which students say they shared with Dias. “I was very, very concerned that one of the probes touching the sample was broken,” one student says. “We could be measuring something that looks like a superconducting drop, but be fooling ourselves.” Although students did see resistance drops in a few other samples, there was no consistency across samples, or even for repeated measurements of a single sample, they told Nature’s news team.
Students were also worried about the accuracy of other measurements. During elemental analysis of a sample, they detected trace amounts of nitrogen. Dias concluded that the samples included the element — and the resulting paper refers to nitrogen-doped lutetium hydride. But further analysis, performed after the paper was submitted, indicated that nitrogen was not incorporated into the LuH. “Ranga ignored what I was saying,” one student says.
Because they were not consulted on the CSH paper, the students say they wanted to make sure they were included in the process of writing the LuH paper. According to the students, Dias initially agreed to involve them. “Then, one day, he sends us an e-mail and says, ‘Here’s the paper. I’m gonna submit it,’” one student says.
As a graduate student, it's a painful process to come to doubt the integrity of the work coming from inside your own group, and especially to start to question the behavior of the professor running the whole effort. You really, really don't want this to be true on either a personal or a professional level, so there are definitely intellectual and emotional barriers that can keep you from coming to such conclusions. You're at the very beginning of your career as a scientist, and you can't help but be overwhelmed by feelings that said career might be getting derailed right at the start through no real fault of your own.
Graduate students are in a particularly vulnerable position here. Different research groups run in different ways, but in almost all of them, the PI's word is at some point equivalent to law. If the professor says "I'm sending in the manuscript" or "I'm taking you off this project", there is very little recourse because of the huge asymmetry of power involved. In many departments there is some theoretical degree of oversight from the department as a whole or the graduate studies office itself, but my impression is that in practice these things are not very effective. For the most part, universities let professors run their research groups in the way that they see fit to, and there's a huge institutional energy barrier to doing anything about it from the outside.
That certainly seems to have been the case at Rochester. As you'll see by reading that article, the university managed to run three separate investigations into the Dias lab and its publications without managing to come to any conclusions of misconduct. That's impressive. The students who spoke (anonymously) to the Nature reporters say that they were never contacted during any of these efforts and were not even aware that some of them were taking place at all. Universities as well are very reluctant to face the facts in these situations, for their own reasons. A fourth investigation, which (finally) brought in outside experts, seems to have resulted in Dias being stripped of his students and his lab space.
The refereeing process for the papers that did get published comes in for some scrutiny as well. There seem to have been very well-founded reasons to doubt the data in the Dias papers, which showed statistical signs of fabrication, but somehow this stuff got published anyway. The Nature article, in the end, is not quite able to explain why Nature published the second Dias paper after the first one had come under fire and after it seems that the great majority of the reviewer reports flagged serious problems with it. Not too many people come out of this story looking good.
The story gets ugly indeed when the students start recording their meetings with Dias, recordings that show that he was trying to manipulate them about the details of Nature's post-publication concerns about the paper. His phrase "We can pretend we’re going to cooperate and buy time for a month or so" is particularly striking, especially as he was removing the dates from the Nature correspondence that he was showing them. In the end, it was the graduate student co-authors who formally requested that the paper be retracted, with Dias himself trying to block them with cease-and-desist paperwork.
I really have to take my hat off to the students involved. This was (is) a horrible situation, and it took a great deal of courage for them to extricate themselves from it. But it should go without saying that things never should have come to this. Of course, Dias himself should not have tried to build a career out of faked data - that should go without saying. But there are always going to be such people, and the question is how far they're going to get before being found out (and how many other lives they're going to mess with before that happens). Rochester's three ineffectual investigations should serve as a strong warning to other institutions and departments. Too often, the purpose of such things seems to be to make the problem go away rather than to find out what is actually happening. Graduate students - human beings in general - shouldn't be treated like this.
The FDA has just approved the first drug (sotatercept) to target the real mechanism behind pulmonary arterial hypertension, and that’s very good news. PAH is not a common disease, but it’s very bad news to get: the huge number of capillaries in the lungs start to narrow, increasing the strain on the heart and causing difficulty breathing, and more than 40% of patients are dead within five years of their diagnosis. The right side of the heart has increasing trouble forcing blood through the narrowing vessels, and the left side then has more trouble supplying the rest of the body. And up until now, there hasn’t been too much that anyone could do about it. PDE5 inhibitors have been one of the main choices, but this treatment and the others available (such as endothelin antagonists) are more about trying to slow the progression of the disease rather than doing anything about its cause. Even with treatment, median survival is still about six years.
That cause seems to be out-of-control proliferative remodeling of the arterial vessels, leading to thicker and stiffer vessel walls with narrower channels. A great deal of research has gone into this process, since it bears on growth-factor signaling in general, and several proteins of the TGF-beta family turn out to be important. These together push a shift towards greater proliferation and lowered apoptosis in the vessel walls, leading to more and more tissue buildup.
Sotatercept itself is a direct intervention in this process. It’s a fusion protein, combining the Fc domain of human IgG1 (which gives you good tolerability and stability in the blood) with the ligand-binding domain of the activin IIA receptor protein. That’s one of the TGF-beta family, and it binds several of the growth factors tightly. Introducing it in this way (where it’s just floating around uncoupled to any further signaling) causes it to be a sponge for the growth factors themselves - you just send the drug in to soak them up.
Now that sounds like a pretty straightforward approach, but very little is ever straightforward in this business. This was the first drug candidate in development at Acceleron back in 2006, and it finally finished a successful Phase III in 2022. The company went public in 2013, at which time sotatercept (and other candidates) were in Phase II trials, and Merck bought the entire company in 2021 for $11.5 billion. Part of the reason this took so long is that those various growth factors are involved in so many different pathways in the body - in fact, the drug has also been considered as a treatment for anemia, for bone growth disorders, for myelofibrosis, and other conditions. This also means that you have to watch carefully for side effects and toxicity. Bleeding and thrombocytopenia were noted in the clinical trials, for example. And as with any protein-based drug, you have to look out for immune-based effects. About a quarter of the patients treated developed some level of antibody response to sotatercept, but fortunately this doesn’t seem to have been associated with adverse events or lowered efficacy.
That efficacy was very notable indeed: both the control group and the treatment group did a six-minute walk test as the primary endpoint of the trial, and the median change in distance covered after 24 weeks of treatment was one meter for the controls and over 40 meters for the treatment group. This is with injections of the drug every three weeks. A list of secondary endpoints improved as well, but that top-line result speaks for itself. Keep in mind that most of the people in this trial were already receiving pretty much every treatment available for PAH, so this improvement is unequivocally beyond anything available with known therapies.
Now, physicians are going to have to keep an eye on hemoglobin and platelet levels as they administer this drug, especially in the first few weeks, and the increased risk of bleeding really is a thing to watch as well. But that level of improvement in a disease that is otherwise a relentless death sentence is worth these sorts of side effects! Congratulations to all involved.
I was in a mood recently. A mood to rummage through hard disks to look at old photos.
I do this every once in a while, stumbling across images I like and wondering, "Hmmm, why didn't I ever process that?"
This usually leads to a bit of futzing, after which I either get distracted or grow weary of fiddling, leaving the photos to be re-discovered another day.
This time I pushed through.
One of the photos I dug up was this image of a pair of sperm whales (Physeter macrocephalus), with the nearer of the two just commencing her dive to forage. I know the whale is a female because this angle provides a clear view of the distinguishing bits, specifically the whale's mammary slits, which is probably why I took the photo in the first place.
Look at the lower ventral area (belly), just forward of where the whale's powerful caudal area (rear end) starts.
See the three grooves?
The two smaller ones on either side of the longer middle one are the mammary slits—charging stations for baby sperm whales.
There are other considerations that provide clues to a given whale's gender, but the only sure-fire way to know is to see this area.
Speaking of which, here is another image I came across—a young humpback whale (Megaptera novaeangliae) playing while her mother chills in the background.
I once heard an "expert" (picture me using fingers to make exaggerated air quotes) assert to a group of people that it is not possible to distinguish the sex of a young whale until it's grown to a certain (unspecified) size (that makes it an adult).
Uh huh.
The mammary slits are present from birth. Provided you can see the correct area, ascertaining female or not is easy.
This little one was having a blast. Seen here, she had just used her fluke (tail) to spin in a circle, pivoting around her upper body region. Mommy was relaxed, conserving her energy for the months of child-rearing that still lay ahead.
Last was this photo of a Bryde's whale (Balaenoptera edeni).
I took this photo waaaaaay back in 2009. We had spotted a pair of these rorquals traveling at high speed in open ocean.
"How fast?" you might wonder.
I'm not exactly sure, but we kicked up quite a wake keeping up. Online references say this species have been clocked up to 13 knots (24kph, 15mph). I'd guess our pair were going a constant five to 10 knots for the more than an hour that we followed. The whales demonstrated curiosity. They pulled up beside our boat two or three times (at full speed of course), giving us a clear look.
My friend Julia and I attempted a single entry. That is all we could do because the whales had taken us far out to sea. We had to go home.
As soon as we hit the water, I shot down to get below the froth from the boat. One whale spun around like Kramer (from the sitcom Seinfeld). I took this photo. The whale shot off in much less than the blink of an eye.
This image has been on my mind ever since.
There was an issue at the time that made it difficult to process the file in a way that would look decent.
We were in open ocean. I was probably 10-15m down. Though it was a clear day, the sun was low. The sharp angle meant diminished light underwater.
The way most digital camera sensors work, you have half the pixels sensitive to green light, a quarter to red, a quarter to blue.
The situation that day meant basically no red light, minimal if any green, and substantially reduced blue due to the low angle of the sun. So out of the 21 Megapixels available to record light (Canon 5DMkii), only a small fraction received stimulation, and those only at suboptimal levels.
That makes for noise (file yuckiness).
Software has improved a lot in 15 years. When I gave the file another look in the latest version of Lightroom, voilà! I was able to reduce the noise enough to make for a decent image. Finally.
More interesting however are the pock marks all over the whale. See them?
Those are the result of cookie cutter sharks (Isistius brasiliensis).
These small sharks live several kilometers (a long way, for you mile-people) down in the ocean. They wait for nightfall to migrate up in search of fine dining. Once cookie cutters find their targets, these sharks (that you may have never heard of before) latch on with their jaws and liberate a chunk of flesh by twisting their bodies. Chomp-twist-extract-swallow. In that order.
Like mosquitoes. Except in the ocean. With teeth. And carnivorous.
I'd like to tell you that you don't have anything to worry about.
OK. I'll do that for the sake of your peace of mind, but whatever you do, don't search for "cookie cutter shark attack human" and definitely don't look at any of the pictures.
The person who is the head of the Republican Party during Donald Trump’s time in office and during his effort to throw out the election result and stay in power any way, and during his effort to run for election again after having done that, is Ronna Romney McDaniel. And she pitched in and helped. She helped set in motion the part of the plot that involved sending fake Trump electors to Congress from states that Trump did not win so Republicans in Washington could use those fake, fraudulent elector slates to contend that maybe Trump did win those states, even though he didn’t.
And don’t believe me on that. There she is on page 23 and page 27 of the federal indictment charging Donald Trump with conspiring to defraud the United States. There is her personal appearance in this scene of the crime as alleged by the U.S. Justice Department in this ongoing criminal case. In Michigan, where the fake electors are themselves now on trial, she told the state of Michigan in writing explicitly, do not certify the election results. The Detroit News has reported that with Donald Trump on the phone with her, she directed Michigan election officials to not certify the vote. She told them, quote, “Do not sign it. We will get you lawyers.”
”