Cancer Initiation: An example of a 6th generation medical science study

Saturday, January 28, 2017

An example of a 6th generation medical science study


Intoduction

 In previous posts, I enumerated the history of medical science paradigms from my point of view. I went on to describe a forthcoming sixth generation.
The underlying sixth generation model consists of a single carbon metabolism cycle that produces glutathione, a convenient biomarker for these studies. Glutathione then is the primary support for DNA methylation by the DNA methyl-transferases, DNMTs.  Study of failure at this level is what is known as  epigenetics study.
 Many disease conditions are now known to be epigenetic in nature. A short list is cancer, lupus, Parkinson's and coronary artery disease.
  An example of what a study of various interacting factors relating to single carbon metabolism is given in Naushad [1].

Whats in that Title?


 The title of Naushad[1] will seem a little unusual so I will discuss just the title for a moment. First of all "multifactor dimensionality reduction" is a term that may be unfamiliar to us old timers. Of course, we are much more familiar with terms like multiple regression and principle component analysis. These "old fashioned" approaches assume that all of the incoming variables are "independent". But what of we don't have a problem that consists of independent variables? In biomedical applications, such as gene/gene interactions or gene/environment interactions, we are looking for a case where the incoming variables are NOT independent. When we find a combination of variables that provides some "zing", then we can start building our model. On other words, we are moving from the world of linear models, to the world of nonlinear models. Lets think of a slot machine. Most of the time, we get a combination that produces no outcome. When the same variables come up "three cherries", we get a big pay out, much bigger that just 3X one cherry, which it typically a zero payout.
   The next term we want to look at is "crosstalk between one-carbon and xeniobiotic metabolic pathways". What this means is that we are going to look at a number of genetic anomalies that may occur in genes that are related to the one-carbon pathway, such as the gene MTHFR, a primary component of the pathway cycling glutathione, 
  The final term is "multi-disease models". In this case, this study is not related to a single medical condition, but four medical conditions,  cancer, lupus, Parkinson's and coronary artery disease, all known to be related to the single-carbon metabolic pathway.

Results

This model that the authors made showed good predictability Parkinson's Disease (PD) and Lupus (SLE). and moderate predictability for breast cancer and Coronary Artery Disease (CAD).
These interaction models showed good predictability of risk for PD (The area under the receiver operating characteristic curve (C) = 0.83) and SLE (C = 0.73); and moderate predictability of risk for breast cancer (C = 0.64) and CAD (C = 0.63).
                                                   Naushad [1] 

Telling us what we know!


Health professionals tell us that we must get a good supply  of vitamins and exercise,   all  of which support good single carbon metabolism. Here we say that genetic components that affect single carbon metabolism are also significant. Good call!.

Direction and suggestion. This type of study and model could be easily extended. For example,  the (non) independent variables (model inputs) could be extended to include known toxins, such as mercury and known nutritional supports, such as folate and vitamin b12. Likewise, dependent variables, or model outputs could autism and Alzheimers.

The immune system (Missing)

 

 So, amid all the excitement about Naushad ( woohoo ), there is an aspect that is missing. All of the clinical conditions enumerated as dependent variables have immunological components. In the case of cancer, the immunological component is thought to be a natural defense. In the others, such as systemic lupus  erythematosus,  the immune response is part of the disease.
A true 6th generation study, as I have defined them, would also quantify the expression of wingspans antigens, as well as the degree of hypomethylation of the patients DNA.

Conclusion

 Biological systems are very different in nature than economic systems, or social science systems. Our system of statistics has grown up around problems where the "independent variables" are um, "independent". In biomedical systems, models have to built that can handle nonlinear interactions between variables. Presumably, a traditional "principle components" analysis would have been insufficient to model this data.
 

[1]  Shaik Mohammad Naushad Sana Venkata VijayalakshmiYedluri RupasreeNadella
 Kumudini Sampathkumar Sowganthika Janardhanan Venketlakshmi NaiduM. Janaki RamaiahDunna
 Nageswara raoVijay Kumar Kutala   Multifactor dimensionality reduction analysis to elucidate the cross-talk  between one-carbon and xenobiotic metabolic pathways in multi-disease models
Molecular Biology ReportsJuly 2015, Volume 42, Issue 7, pp 1211–1224  [Abstract]






No comments:

Post a Comment