<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"><html>
head><meta http-equiv="content-type" content="text/html;
charset=utf-8"></head><body><div dir="ltr"><p>Dear Alberto,</p><p>
</p><p>Many thanks for the kickoff text. I will try to produce a
couple of direct comments.</p><p>You have reminded me of the early
70's, when I first approached science. A few computers had made their
entrance in the university halls. During those years, and for some
decades to come, a new mantra was to be ensconced: modeling,
simulations. Thanks to computers, we had a fascinating new tool; a
mathematical machine that was opening a new window to the world of
science, equivalent to the telescope or the microscope in the
scientific revolution. Now, almost 50 years later, after having
provoked their own "information revolution" it seems that computers
are more than a new tool. Dataism coupled with artificial
intelligence, deep learning and the other techniques, have taken them
to the command post, so that they are becoming direct "agents" of the
scientific progress. And this is strange. They have already defeated
masters of chess, of go and of other contests... are they going to
defeat scientists too? Are they the "necessary" new lords of all
quarters of techno-social complexity?</p><p>You have depicted very
cogently the new panorama of biomedical research, probably the
mainstream, and I wonder whether this is the most interesting
direction of advancement. In some sense, yes (or no!), as it is where
big biomed companies, technological firms, and management
establishment are pointing at. It is easy to complain that they are
leaving aside the integrative vision, the meaningful synthesis that
facilitate our comprehension, the "soul" in the machine... But we have
been complaining in this way at least during the last two decades. So
I really do not know. Fashions in science come and go: maybe all of
this is a temporary illusion. Or a taste of the science of the future.
</p><p>In any case, it was nice hearing from a biomedical researcher in
the wet lab.</p><p>Best wishes--Pedro</p><p> </p><p>On Tue, 06
Mar 2018 21:23:01 +0100
"Alberto J. Schuhmacher" wrote:</p>
blockquote><p>Dear FIS Colleagues,</p><p>I very much appreciate this
opportunity to discuss with all of you.</p><p>My mentors and science
teachers taught me that Science had a method, rules and procedures
that should be followed and pursued rigorously and with perseverance.
The scientific research needed to be preceded by one or several
hypotheses that should be subjected to validation or refutation
through experiments designed and carried out in a laboratory. The
Oxford Dictionaries Online defines the scientific method as "a method
or procedure that has characterized natural science since the 17th
century, consisting in systematic observation, measurement, and
experiment, and the formulation, testing, and modification of
hypotheses". Experiments are a procedure designed to test hypotheses.
Experiments are an important tool of the scientific method.</p><p>In
our case, molecular, personalized and precision medicine aims to
anticipate the future development of diseases in a specific individual
through molecular markers registered in the genome, variome,
metagenome, metabolome or in any of the multiple "omes" that make up
the present "omics" language of current Biology.</p><p>The
possibilities of applying these methodologies to the prevention and
treatment of diseases have increased exponentially with the rise of a
new religion, <em>Dataism</em>, whose foundations are inspired by
scientific agnosticism, a way of thinking that seems classical but
applied to research, it hides a profound revolution.</p><p>Dataism
arises from the recent human desire to collect and analyze data, data
and more data, data of everything and data for everything-from the
most banal social issues to those that decide the rhythms of life and
death. “Information flow” is one the “supreme values” of this
religion. The next floods will be of data as we can see just looking
at any electronic window.</p><p>The recent development of gigantic
clinical and biological databases, and the concomitant progress of the
computational capacity to handle and analyze these growing tides of
information represent the best substrate for the progress of Dataism,
which in turn has managed to provide a solid content material to an
always-evanescent scientific agnosticism.</p><p>On many occasions the
establishment of correlative observations seems to be sufficient to
infer about the relevance of a certain factor in the development of
some human pathologies. It seems that we are heading towards a path in
which research, instead of being driven by hypotheses confirmed
experimentally, in the near future experimental hypotheses themselves
will arise from the observation of data of previously performed
experiments. Are we facing the end of the wet lab? Is Dataism the end
of classical hypothesis-driven research (and the beginning of
data-correlation-driven research)?</p><p>Deep learning is based on
learning data representations, as opposed to task-specific algorithms.
Learning can be supervised, semi-supervised or unsupervised. Deep
learning models are loosely related to information processing and
communication patterns in a biological nervous system, such as neural
coding that attempts to define a relationship between various stimuli
and associated neuronal responses in the brain. Deep learning
architectures such as deep neural networks, deep belief networks and
recurrent neural networks have been applied to fields including
computer vision, audio recognition, speech recognition, machine
translation, natural language processing, social network filtering,
bioinformatics and drug design, where they have produced results
comparable to and in some cases superior to human experts. Will be
data-correlation-driven research the new scientific method for
unsupervised deep learning machines<em>? </em>Will computers became
fundamentalists of <em>Dataism</em>?</p><p>Best regards,</p><p>AJ</p>
p> </p><p>---</p><div class="pre" style="margin: 0; padding: 0;
font-family: monospace"><span style="color: #000000; font-family:
arial,helvetica,sans-serif; font-size: 10pt;">Alberto J. Schuhmacher,
PhD.</span><br><span style="color: #000000; font-family:
arial,helvetica,sans-serif; font-size: 10pt;"> Head, Molecular
Oncology Group</span><br> <br><span style="color: #000000;
font-family: arial,helvetica,sans-serif; font-size: 10pt;"> Aragon
Health Research Institute (IIS Aragón)</span><br><span style="color:
#000000; font-family: arial,helvetica,sans-serif; font-size: 10pt;">
Biomedical Research Center of Aragon (CIBA)</span><br><span
style="color: #000000; font-family: arial,helvetica,sans-serif;
font-size: 10pt;"> Avda. Juan Bosco 13, 50009 Zaragoza (Spain)</span>
br><span style="color: #000000; font-family:
arial,helvetica,sans-serif; font-size: 10pt;"> email: <a style="color:
#000000;" href="https://mailto:ajimenez@iisaragon.es" target="_blank">
ajimenez@iisaragon.es</a></span><br><span style="font-family:
arial,helvetica,sans-serif; font-size: 10pt;"> Phone:(+34) 637939901
</span><p> </p></div></blockquote><p> </p></div></body></html>