Modeling life beyond Metabolomics. A comprehensive approach to metabolism in health and disease.
What models can predict
The role of Models in Clinical Practice
The EBM approach
- Symptom/disease identification
- "correct" therapy on the basis of the evidences (average choice based on thousand cases)
The biochemical approach
- Identification of biochemical causal relationships leading to symptoms/disease in the single individual
A network neuroscience of neurofeedback for clinical translation, 2017
Patient is unique (read: anecdotal) so why he should be treated as similar to ....?
“Anecdotal Evidence”: Why Narratives Matter to Medical Practice, 2006
When anecdotes fit a deep biochemical model they are no longer anecdotes but evidences of biochemical model fitness.
Biochemical network are made by physical agents that are always linked by causal relationship, even if you don't know them.
When the network is very complex to test the predictive power of simple hypothesis on a large number of subjects usually fails because of the large number of factor involved but not recorded. ICT allows at acceptable costs collection of huge amount of data.
When an anecdotal info, collected by chance, lighten a new causal node, this node is added to the network and checked for other instances. And so one until the vast majority of the events become predictable.
Evidence is a record of the past, not a preview of the future.
The only correct experimental evidence is the never ended one, and evolutionary like life.
A Biochemical Method can allow Previews: if they are confirmed in the real patient the model is reinforced, otherwise we are forced to improve it (new nodes, new relationships, new polymorphysms)
‘Deep’ models and their relation to diagnosis, 1989
We provide a typology of different knowledge structures and reasoning processes that play a role in qualitative or functional reasoning and elaborate on functional representations as deep cognitive models for some aspects of causal reasoning in medicine.
The Failure of AI in Medicine depends on the fact that it is based on deep cognitive models and not on deep biochemical models
What is the Disease?
Disease: inadequacy of a living organism to cope with environmental requirements
To understand Disease we have to define the Living Organism and the Rules driving its life
Example 1: Pain
The First Rule: Dissipative
the higher the stimulus, the higher the spikes frequency...
but we get the same results when the intracellula ATP is low....
which local conditions can affect ATP synthesis?
- Less oxygen transport
- Less CoQ
- Less Cyt C
- low estrogens--> low heme synthesis
- iron deficiency
- Less glucose
Pain, the First Rule and the Fourth Rule
Example 2: Regulation of proteins synthesis by Amino Acids Availability
Aminoacids supply to the tissues
livello di O2 e attivita' cognitive
Brain metabolite levels and language abilities in preschool children, 2015
viene fatta la rilevazione dei metaboli mediante la "Magnetic resonance spectroscopy" (MRS) e l'LCModel (Provencher, 2001) per calcolare la concentrazione (mmol/kg ww) per:
glutamate (Glu), glutamine (Gln), creatine + phosphocreatine (Cr/PCr), choline (glycerophosphocholine
+ phosphocholine), inositol, and NAA
in relazione alle abilità di lettura in bambini in età prescolare
"Our results show for the first time that brain metabolite concentrations and language abilities are related in preschool children.
Specifically, phonological processing ability was positively associated with glutamate, creatine, and inositol concentrations in the anterior cingulate"
A pag 6 dell'articolo ci sono i grafici relativi alla correlazione tra la concentrazione dei metaboliti e la velocità di lettura.
Alte concentrazioni di inositolo, creatina e glutammato nel giro cingolato anteriore sono correlati a buone abilità di lettura nei bambini.
La colina e la glutamina sembrano invece essere inversamente correlati con le abilità di lettura, in particolare la glutamina.