Month: February 2014

Minimal volume with minimal information

The WIRED interview with Elizabeth Holmes, founder and CEO of Theranos, has been making the rounds. In Di Cleverly’s post

https://plus.google.com/+DiCleverly/posts/XTn2SDnE9LB

I mentioned that in addition to the minimal amount of blood used, there is a minimal amount of details. Searching elsewhere, even the Theranos website, there aren’t many details. In Di Cleverly’s post, the only decent info was from patents. If you’ve ever read a patent, then you know that it’s often difficult to sort out what’s really going on. So, with Di Cleverly’s help, we have a better picture of what’s going on. A lot of this post are my guesses about some of the details, partly because I’m busy, partly because I’m lazy, and partly because there isn’t a lot out there without really digging. Did I mention I’m lazy, I mean busy?

☼ What was mentioned: small volume and centralized facility

For those that haven’t seen the WIRED or Medscape articles, Elizabeth Holmes dropped out of college at Stanford at the age of 19 and eventually started Theranos with her college funds. The interview talks about how the small volume of blood, from a pin prick, can make the experience, and therefore patient compliance, better. Ms. Holmes talks about reduced and transparent pricing. Essentially none of the technology is discussed. A centralized facility is mentioned. So is that an essential part, i.e., how much can be done off site (e.g. at Walgreens)? Before any young readers decide to drop out like Ms. Holmes or Bill Gates, I think Dave Thomas, founder of Wendy’s makes a good example.

Thomas, realizing that his success as a high school dropout might convince other teenagers to quit school (something he later claimed was a mistake), became a student at Coconut Creek High School. He earned a GED in 1993.)

http://en.wikipedia.org/wiki/Dave_Thomas_(businessman

☼ Detective Work: ESR and microfluidics

On Di Cleverly’s post some detective work was done and a few things came to light, mostly via the patents. The small “nanotainer” is used in a novel centrifuge to get information about the blood sample. Red blood cells (RBC) are called erythrocytes and are just one component of blood. If you put whole blood in a glass tube, eventually the RBCs will sink to the bottom and the plasma will stay at the top. You can speed up this process by using a centrifuge (a device that spins the tubes at many times the force of gravity). The rate that the RBCs go to the bottom is called the erythrocyte sedimentation rate or ESR. ESR alone can tell you something about your health.

An increased ESR rate may be due to:

Anemia

Cancers such as lymphoma or multiple myeloma

Kidney disease

Pregnancy

Thyroid disease

Common autoimmune disorders include:

Lupus

Rheumatoid arthritis in adults or children

Very high ESR levels occur with less common autoimmune disorders, including:

Allergic vasculitis

Giant cell arteritis

Hyperfibrinogenemia (increased fibrinogen levels in the blood)

Macroglobulinemia – primary

Necrotizing vasculitis

Polymyalgia rheumatica

An increased ESR rate may be due to some infections, including:

Body-wide (systemic) infection

Bone infections

Infection of the heart or heart valves

Rheumatic fever

Severe skin infections, such as erysipelas

Tuberculosis

Lower-than-normal levels occur with:

Congestive heart failure

Hyperviscosity

Hypofibrinogenemia (decreased fibrinogen levels)

Low plasma protein (due to liver or kidney disease)

Polycythemia

Sickle cell anemia

Source: http://goo.gl/zKstuW 

The patent mentions a novel centrifuge device with either video or still images of the sample. There are two greyscale figures from the patent in the album below. With image analysis the ESR can be measured without human intervention which minimizes errors.

☼ Microfluidics

Another patent talks about microfluidic devices. I’m assuming those are lab-on-a-chip (LOC) devices. LOCs use microelectromechanical systems (MEMS) to do analysis on very small volumes of fluid. Here’s an example from Harvard that captures trace amounts of tumor cells.

http://goo.gl/jPGZlr

Although genechips or DNA microarray’s aren’t LOCs, it is possible they are being used by Theranos. An image of an Affymetrix Genechip is included in the album below. http://goo.gl/GpMyjx Note the small Eppendorf tubes in the foreground. Those are larger than the Theranos “nanotainer” but they do make Eppendorf tubes the same size as the “nanotainer”. Both the “nanotainer” and Eppendorf tubes have conical bottoms to facilitate removal of all of the liquid. The genechips have target DNA probes attached to the device. If a target gene is expressed, it will bind with the probe on the chip. The readout is typically some type of light whether chemiluminescence, fluorescence, or some combination. The amount of information from these genechips has caused an explosion in bioinformatics and computer processing dedicated to speeding up the analysis of these microarrays.

http://en.wikipedia.org/wiki/DNA_microarray

Because the samples are going to a centralized facility, it’s possible that real-time polymerase chain reaction (RT-PCR) is also being used. RT-PCR is a technique that is used to amplify DNA samples.

☼ Therapeutics and Diagnostics =  Theranostics

I didn’t find any information to suggest that the name Theranos has anything to do with the term theranostics, i.e, therapeutics and diagnostics.

Pharmacogenomics aims to identify the genetic basis of variability in drug efficacy and safety, and ultimately develop diagnostics that can individualize pharmacotherapy. Theragnostics, a term denoting the fusion of therapeutics and diagnostics, is receiving increasing attention as pharmacogenomics moves to applications at point of patient care.

Shifting emphasis from pharmacogenomics to theragnostics

http://goo.gl/6nrkmp

Rapid molecular theranostics in infectious diseases.

Picard FJ1, Bergeron MG.

Drug Discov Today. 2002 Nov 1;7(21):1092-101.

http://www.ncbi.nlm.nih.gov/pubmed/12546841

An example of theranostics from my boss and colleagues is a platform that combines doxorubicin (cancer therapy), herceptin (targeting for diagnosis), and DOTA-Gd(III) (for MRI detection, i.e, diagnosis). So the herceptin targets the product to cancer cells. Gadolinium, chelated to the construct (DOTA-Gd(III)) allows you to see it with MRI (enhances the contrast from background tissue) and the doxorubicin provides therapy at the target (tumor).

pH-Responsive Theranostic Polymer-Caged Nanobins: Enhanced Cytotoxicity and T1 MRI Contrast by Her2 Targeting

http://goo.gl/4PHkLo

So that’s what I could sort out with the help of Di Cleverly’s post and my own digging through a couple patents. If you have ideas or comments, feel free to ask.

#ScienceSunday  

Impact

During a HOA there was a discussion about Open Access journals. Brent Neal wrote a follow up post

Open Access

http://goo.gl/PbQaO0

where I mentioned that journal impact factors should be discussed as they play a role in accessing the quality of a journal. Some Open Access journals are good and some are not so much. How can you tell? There’s some nuance and disagreement about impact factors but I’ll get to that later.

First, I want to give a little background and continue the conversation about Open Access journals. In Brent’s post, he mentioned predatory publishers and that we have all gotten spam from them, i.e., requests to consider Open Access journal X when we publish our next manuscript. One of the negative sides of predatory Open Access that I’ve experienced is related to peer review and the role of the editor. After you have done your job reviewing a manuscript and recommended whether or not the manuscript should be accepted for publication, sent back for major revision or rejected outright, the editor takes into consideration the recommendations from all referees and informs the author(s) of his/her decision. The problem is that some predatory Open Access journals charge a significant amount to the authors, sometimes more than $1,500. In the case that I am thinking of, the manuscript was poorly written and was essentially what is known as a quick communication that was being submitted as a full research article. The manuscript was very verbose to try justify full article vs. quick communication. The editor kept pushing to accept the article and to accept it as a full research article. I can only guess the motivation for that was the fee that is charged to the author(s).

Removing the issue of predatory journals, how does one assess the quality of a journal and more importantly a specific journal article. I’ll discuss an example. You probably hear scientists on G+ request peer reviewed citation when “debating” with people. I put debating in quotes because people often don’t know what it really means, it does not mean arguing but I’ll save that for another post. In a debate with a commenter on one of my posts (sorry I couldn’t find the comment to link), he finally gave a link to a peer reviewed article in Bulletins in Insectology. I’m not an entomologist so I have no idea of the accuracy or impact of that particular article. So what do you do?

Phone a friend

Like the Who Wants to be a Millionaire show, one option is to ask an expert. Maybe you know an entomologist. One of the great things about G+ is that you might actually have one in your circles. Alas, I don’t know any or at least couldn’t think of one. The next option is to assess the quality of the journal using impact factors.

Impact Factor (IF)

Journal Citation Reports are made by Institute for Scientific Information (ISI) and can be found on the ISI Web of Knowledge site, owned by Thomson Reuters. You have to have a subscription to the site so this ties in with the Open Access discussion from an access point of view as well. Impact factor is a calculation of the average number of citations per paper for a 2 year period divided by the total number of citable items. The idea of IF is that it gives you an idea of the average importance or impact of articles for a journal. You can imagine where there can be problems with this, e.g., what if a journal publishes a low number of articles per year? The Wiki below goes through more explanations and some alternatives like Page Ranking. A good example in the Wiki is an article that was cited over 6,000 times, yet the other citations for that journal are much lower.

Getting back to the Bulletins in Insectology journal, I looked it up in the Citations Reports. It’s impact factor is 0.44. In the post with the “debate” I had referenced an article in the Proceedings of the National Academy of Science (PNAS). It’s impact factor is 9.737. Just for reference, Science has an IF of 31.027 and Nature has an IF of 38.597. You’ll see them along with other details in the citation report, in the attached figure. Here’s the problem, most people agree that you can’t really compare IF from different disciplines. One reason is that some research might take longer to complete and publish. So if one discipline churns out more publications, that will affect the IF. The number of articles from Bulletins in Insectology is only 42. Remember, IF divides by the number of citable items and a low number should help.

Entomology is a reasonable category to compare Bulletins in Insectology  with other journals, in the same discipline. In the next figure below, you’ll see a screenshot of the first page of results (sorted by IF) for journals in entomology. The range of IF is from 13.589 to 1.926. Without being able to “phone a friend”, one would conclude that Bulletins in Insectology is either an obscure journal, new journal, or one that is not ranked high in entomology.

So the Bulletins in Insectology article that was linked is peer reviewed, which is good, but it likely has some issues preventing it from being published in a better journal in the field of entomology.

In our own fields of research, we often don’t pay too much attention to IF because we know which journals our peers/colleagues are publishing in. Unfortunately some academic administrations will use impact factors to judge the quality of someone’s track record for promotion purposes. Again, if they are not in your field, it is one way to assess quality, albeit with the caveats mentioned.

http://en.wikipedia.org/wiki/Impact_factor

Homer GIF via Reddit

#ScienceSunday  

Peer review causes humility

If only the average Joe or Jane could experience having their manuscript ripped to shreds during peer review. Sometimes it’s legitimate and sometimes it’s just a referee that woke up on the wrong side of the bed. Either way, the author has to suck it up and genuflect.

Brian Koberlein explains why science is humbling.

#ScienceSunday  

Originally shared by Brian Koberlein

Humility

Yesterday’s post about the big bang and cosmic origins struck a few nerves.  Responses ranged from vulgar insults to dismissals of the post as “just a theory.”  But more subtle were the criticisms that declared the post lacked humility.  Scientific knowledge is never perfect, and to claim the validity of the big bang is to go too far.  When communicating to the general public scientists should never say “we know”, only that “we might know.” Scientists should show more humility.

Such criticism fails to recognize that the power of science is its humility.  In fact, the scientific process is based on the assumption that individual scientists won’t easily show humility on their own, so it is imposed upon them. There are three basic tenets of scientific research: it must be based upon verifiable data, it must be done publicly, and it must be open to criticism.

Most people view scientific evidence as repeatable experiments that can be done in the lab.  For this reason the findings of evolution or cosmology are often countered with “you weren’t there.”  But verifiable data is much broader than simply lab experiments.  It is a process of gathering data that clearly documents when, where and how the data was gathered.  If you gather observational data, the burden is on you to document its origin.  If you use data gathered by others, you must clearly cite your sources.

Once you have your observational results or theoretical work, the next step is to present it publicly.  This could be a conference, a preprint archive, a book, or submission to a research journal.  A scientific discovery is meaningless if it isn’t disseminated.  Publication provides a record of the work, so it can’t be tossed down the memory hole.  Make a significant discovery, and the record is there.  Make a foolish claim, and that’s there too.  It’s the latter possibility that strikes fear into scientists everywhere, because  publishing your work isn’t sufficient.  When you make your research public your colleagues now have a chance to pull the work apart and see if it really says what you think it says.  It gets subjected to peer review.

Peer review can be the most frustrating and most humiliating aspect of scientific research.  That’s why it’s considered the gold standard of science.  Having research published in a peer-reviewed journal means that the work has been examined by other experts in your field, and has been found clear and without obvious error.  It doesn’t mean its perfect, but it does mean the work has been held to a high standard and survived.  This is why when I write about new scientific work I focus on peer reviewed articles.  When I write about work that hasn’t been peer reviewed, I clearly say so.

Of course even after conducting your research, organizing your results, checking it with friendly colleagues, presenting it publicly and submitting it to peer review, you still aren’t done.  You’re never done, because at any time someone can critically review your work again.  If you have a great theory and your predictions don’t support new findings, we look for something better.  No matter how famous, or how many awards you may have, anyone can be toppled by new scientific discovery.

That’s the deal.  Keep pushing back against ideas.  Keep working to develop better theories.  Always, always keep in mind that your theories might just be wrong.

What survives is an understanding of the universe that it robust.  It is a confluence of evidence that supports a deep theoretical framework.  It is knowledge humbly gathered, and put forward with humility.  Through a process that recognizes human fallibility.  It is humanity’s best understanding of what is real and true about the cosmos.

This is why I present ideas like the big bang with the claim that we know.  We Know.  We know because thousands of individuals have devoted their lives to understanding the universe.  Devoted their lives to getting it right.  Relying on a process that forces us to be humble, and forces us to defend our ideas over and over.

In my posts I always strive to present our best understanding of the universe in a way that is clear and meaningful.  That’s why I try to limit moderation of the comments.  It is a kind of peer review.  I write about science to the best of my ability, and everyone is free to criticize it.  I’ve made mistakes in my posts and been called on them.  I’ve been praised and thanked for making things clear.  I’ve also been called a liar. A fool. Prideful. Deceitful. Ignorant. Arrogant.

Fair enough.  That’s the deal.

Image:  Excerpt from da Vinci’s notebooks.