Physics saves humanity

Recent news that a blood test could detect early-stage cancer with a 65% accuracy (1), is promising. However, this is not a sensitivity level that makes such technology very useful. Life sciences and healthcare researchers have been suffering from segmented specialization and domain experts in each sub-segment believe that they know everything. This has led to underutilization of available technologies from other industries and solutions that optimize within a narrow context. If the goal is to reach the best possible solution, it is advisable to get out of the labs and look across domains and let some of the egos go.

Healthcare, perennial laggards in the use of information technology, in the prediction, diagnosis, and treatment of diseases is falling further behind. As the engineers figure out autonomous cars and space tourism without breaking a sweat, life sciences and healthcare professionals, steeped in conventionalism, have been pretending that humans are indeed different from machines. Certainly, the policymakers in Washington appear closer to machines as they "retire," with lifetime healthcare benefits after robbing the same from 13 million Americans. And the most powerful one, after figuring out the 140 character idiot box, has been addicted to it just as a robot would be.

Machines are accelerating toward demonstrating higher cognitive capabilities while the frail bodies of the declining species suffer from a lack of acceptance of change. They have been immensely creative at inception. As they stood up in the African Savannah with a feeble architecture that was no match to the beasts that roamed, they courageously exposed themselves to danger. They traveled to every corner of the blue planet on foot and created habitats that are in sync with the environment. They survived a narrow bottleneck of fewer than 15,000 samples as the ice age advanced across the globe. And then, the "modern woman," arrived - and that was bad news. Agriculture, the industrial revolution, and computer technology seem to have made them weaker. Grains made them diabetic, industries have been fuming poison into their greenhouse and technology now appears to set them back.

The fundamental question remains to be that if life indeed is a result of Physics. Before the "God particle," and "gravitational waves," there were more fundamental concepts such as entropy. If entropy has an unambiguous positive slope and more importantly if there is a universal objective function that maximizes entropy, life certainly fits. Life appears to be most efficient compared to natural processes to accelerate entropy and that points to the idea that the creation, maintenance, and eventual destruction of life are driven by physical processes. To reject this hypothesis, one has to prove that life has entropy reducing effects. It does not appear to be so. Organization of life in structures from bacteria to humans appear to accelerate entropy. It is possible that one can mathematically show that the size of colonies of life that we observe is entropy maximizing.

Physics may require life to survive as it may be the best way to maximize an overall objective function. Humans may be saved in spite of themselves, by Physics.


(1) http://www.sciencemag.org/news/2018/04/blood-test-shows-promise-spotting-early-cancers

The bane of pharmaceutical R&D

A recent study (1) appears to raise red flags on pharmaceutical research, animal studies and the contemporary scientific process in general. Perhaps it could be new to the authors but most of what they describe have been known to the community for many decades. The following are important considerations in this debate. I state them without proof but there is plenty out there:

(a) A very large percentage of the published studies cannot be replicated
(b) Most of the published studies target proving something rather than the other way around
(c) The quantity of publishing (rather than the quality) is the most important metric for most educational institutions to determine the reward for academics
(d) Big pharma is run by outdated leaders who are trying to churn out incremental medicines to meet shareholder value targets
(e) The drug discovery and development processes are ably assisted by an incompetent regulatory agency with many conflicts of interest
(f) Hypothesis testing in life sciences still clings to a nearly 100 years’ old idea that uncertainty is normally distributed. And most statisticians, encumbered by the agency’s love for “p-value,” will not deviate from the framework. And in the process, they have approved bad drugs, rejected good ones and failed to identify sub-populations who could benefit from the NCE.

So, the authors’ contention that many animal studies are not published at all, albeit interesting, is just the tip of the iceberg. There is a much bigger problem to tackle. The leaders of life sciences companies and their regulators may want to consider retirement, say after 80, as they may need to yield to young leaders who have a higher appreciation of emerging technologies.

The correlation between animal studies that precede the clinic and what happens in humans has been incredibly low for almost a century. They have tried everything from mice, rabbits, dogs, and chimps in an attempt to prove the unprovable. In the process, they reduced animal welfare while simultaneously developing therapies that can only be called, “bad.” The finding that the therapeutic index of marketed drugs seems to decline over time is a warning signal that there are many inefficiencies in the R&D and approval processes.

Technology is advancing. We do not have to stick to regression slide rules to prove or disprove if a drug works anymore. It is time life sciences industry embraced ideas that are transforming every other industry. To make that happen, it will require cleaning the shop and starting over.

Old ideas die hard and older ideas are even worse.

(1) http://www.sciencemag.org/news/2018/04/clinical-trials-may-be-based-flimsy-animal-data