Thursday, May 28, 2015

Ad Hoc Propiter Hoc

In a recent piece in Vox the authors state:

Compared to other developed countries, the US ranks high on income inequality and low on social mobility. This could be particularly concerning if such a trend is self-perpetuating. In this column, the authors argue that there is a causal relationship between income inequality and high school dropout rates among disadvantaged youth. In particular, moving from a low-inequality to a high-inequality state increases the likelihood that a male student from a low socioeconomic status drops out of high school by 4.1 percentage points. The lack of opportunity for disadvantaged students, therefore, may be self-perpetuating.

Perhaps they have never been to France or the UK. In the UK it is your birth family that all too often defines you. It defines your position in society and your outcome. In France it again is family and school. In Russia, well we all know that one.

In the US anyone has a chance despite the author's contentions. Moreover one should look at their own data to see that DC has the highest drop out rate the the highest income disparity! Why, the Government. Lobbyists get millions and the old time residents wallow in a failing school system, albeit supported by the taxpayers.

In my experience the US has the greatest social mobility of any country. Anyone can start anything and try to make a go of it. Failing to complete High School is all too often a failure of the local Government and its way to educate. One need just look at the Charter Schools and their success.

Sunday, May 24, 2015

A Book on the Genomics of PCa

Prostate Cancer by Stabiano is a recent contribution to the field of literature discussing the genetic underpinnings of this disease. It is a multi-author work by those at the University of Naples and surrounding institutions. It consists of eighteen chapters and the focus is on describing the recent efforts on various elements of the genetic bases of prostate cancer. Each chapter is prepared by one or more of the authors and covers a somewhat specific topic.

Chapter 1 is a well written summary of pathological methods of current use in diagnosing PCa and its related presentations. The materials is well presented and sets the stage for the discussions regarding the genetic elements.

Chapter 2 sets the tone of the remaining Chapters. This Chapter discusses inflammation. Its style is carried through the remainder of the book. Namely the author presents briefly the importance of inflammation and then proceeds through the most recent literature on a topic by topic basis covered in one to two sentences. The presentation is more akin to a literature survey rather than an introduction or even detailed discussion of inflammation in PCa. Inflammation has always been a concern in PCa and its study has been spotty at best. To understand inflammation and its effects one must understand what genes may be affected as well as the impacts on methylation and miRNAs. This becomes a somewhat circular analysis and the authors set up each separately and on a somewhat standalone basis.

Chapter 3 is on apoptosis. This is a key element in almost and cancer. This is a well done chapter and does try to tie together some of the elements. The especially useful addition is the discussion on non-apoptotic elements which is fairly complete and of significant interest to those seeking an expansion of this work.

Chapter 4 discusses the AR, androgen receptor, dynamics and it again follows the style of reviewing and commenting upon the current literature. Chapter 5 does likewise for neuroendocrine issues.

Chapter 6 is an excellent chapter on metastasis. It covers a great deal of the current work including for example that on ALDH1A1 and potentially ALDH1A3. The graphics are generally good and useful. Chapters 7-11 continue the discussion of pathways and their interactions.

Chapter 12 is a discussion of epigenetic factors including methylation and miRNAs. This is a powerful area of research and it would be useful to have expanded the discussion. Whether or not SNP issues fit here as well is an open debate. However, epigenetic factors are becoming significant in understanding many cancers, since they can change expression while leaving DNA in its original form. Epigenetics blocks mRNA from converting into their operative proteins. This may then become a viable path for a therapeutic.

The remaining chapters cover a wide variety of related topics. All are written in the same manner

Overall the books is an excellent source of accessing the literature. It is, however, neither an introduction for those seeking to understand all of the elements, nor is it a standalone text useful for in depth understanding. I would strongly recommend it for those studying the genetic factors associated with PCa. It is an excellent addition and expands understanding the literature. However it is not for the individual seeking a first exposure and it is not a document which attempts to provide unification of the topic.

I would like to have seen some detailed discussion on the issues associated with the loss of cell fixation by the breakdown of the extra-cellular matrix adhesion. I would also have like to see a more detailed discussion of the pathways and their interactions. So much of the book is a single sentence statement of third party work that the sense of cohesion is oftentimes missing.

Overall, however, this is definitely worth having as a reference source for those in the field.

What's in a Name?

Names can mean something, should mean something. So I was surprised when I read a WAPO piece by a University President proposing some "innovative" way to monetize research.

The article states:

To create a new way of supporting the first stage — from idea to investment — a coalition of funders from the public, for-profit and not-for-profit sectors could work together to establish “innovation orchards.” These would provide what universities alone cannot: the physical space, mentorship and bridge-funding for entrepreneurs to turn new science into workable products, up to the point that they meet venture capital’s five-year threshold for the journey from investment to an impact on the market. This would make investing in tangible or tangible-digital hybrid innovations no riskier than investing in the purely digital. 

It seems to propose that some group or groups "fund" the process of going from pure research to implementation with this "Orchard" concept. Frankly funding of all types seeks a return. The VC or similar funding entity has a lifetime associated with it. It does not go on forever and its returns have a discount factor. You can promise billions of dollars but if it is thousands of years from now one will have little to no interest.

One can suggest something but perhaps more meat on the bones would have been useful. There once was a time when University Presidents could suggest with some detail. This is throwing a name on the wall and hoping that others will fill in the gaps, there being many.

In my experience I a reminded by a comment made by a colleague, oft repeated, "A good idea does not a business make" It is quite unfair to researchers to let them think that the idea alone, even if it is embodied, has any merit other than its very existence. University research is a process of training, teaching a student how to accomplish a task. True research is just that, exploring the future. However one should not expect that every research result is productizeable. They are not.

The real issue is that in our current research the biotech world is soon to dominate. Unlike our past high tech adventures this world will be controlled by the Government, the FDA. It is that process which will delimit what we can accomplish and it is that world which needs modification.

Thursday, May 21, 2015

Dig a Bit Deeper!

In a recent Nutrition and Diabetes journal article the authors conclude:

Most studies examining the relationship of body mass index (BMI) with mortality in diabetic persons suggest a paradox: the BMI category with lowest associated mortality risk (overweight or obese) is higher than it is in non-diabetic persons (normal weight)

Let's try this in English. They seem to state that Type 2 Diabetics with lower BMI have higher mortality.

They continue:

In conclusion, in comparing physical and mental health status by BMI category in concurrent national samples of diabetic and non-diabetic persons, we found evidence of a physical (but not mental) health status paradox in diabetes. Physical health status was most optimal in the overweight category among diabetic persons, versus in the normal-weight category among non-diabetic persons. Given that physical health status influences mortality risk, the findings suggest possible pathways to a BMI mortality paradox in diabetes.

Now this article leaves several questions. They seem to mix Type 1 and Type 2 Diabetes. It is not really clear in reading.

Second, perhaps if one has Diabetes and has low BMI then one may have a totally different form of Diabetes.

Third, in examining anything like this one needs the history of each patient. Not just a single time sample. How heavy were they when the Diabetes started? Just to begin with.

The Tables state Diabetes, but what type?

Type 2 Diabetes with BMI under 20? I find that amazing. One is on the verge of ketosis and yet they have Type 2 Diabetes. It makes no sense. Unless of course it is something else.

Perhaps we should have seen a picture of the authors? Just a thought.

Tuesday, May 19, 2015

Yield Curves May 2015

The Treasury Yield curves are shown above. They seem to have clumped and are now spreading out again
The above is an example show today's curves and lows and highs. The question is: what do these really reflect?
Finally the spreads, starting up again but not a great deal. In many ways this is still very artificial.

Free Tuition

There is a movement to have free tuition at state universities funded by the Feds, aka the taxpayers. Is this a good idea?

Frankly if students who perform well and demonstrate their abilities, then in state tuition at their state schools is great. West Virginia already has it for high performers and New York used to have Regents scholarships before they did away with Regents. It makes sense for those who are doing well but not MIT or Harvard yet.

But one must beware of several factors:

1. It should be performance based. What performance? Class ranking, grades, even a competitive exam. There must be some clear demonstrable level of performance. It must also be blind to anything else. Otherwise it will be gamed.

2. The Feds must be hands off. We all know what the Feds up costs. How can this be done, block grants perhaps.

3. Avoid social engineering. If one wants people to perform then make the rules clear and play the game. We do not social engineer football, basketball, etc.

4. Promote productive studies. We really do not need too many fine art majors or French literature majors. We do need engineers. We don't need more lawyers, we do need competent physicians.

5. Reward performance and punish failure. Yes, if they fail then drop them. Give a second chance but do not let them linger.

6. Set standards, real standards for real life. The goal of education is to be productive. Productive in society. One can always study philosophy later. Or at the same time if one can fit it in.

Just some thoughts.