Machine Learning-assisted Research on CFS

[mariovitali]

Here is a quick update.

First of all, i emailed a 33-page document to @Janet Dafoe (Rose49) that explains the "Multiple Pathways Hypothesis" so that Prof. Ron Davis has a look at it.

Using Machine Learning, the importance of Proteolysis, Ubiquitin-Proteasome System, Protein Degradation and Catalase were found. Also, the potential importance of L-Lysine and Digestive Enzymes. I will have to say here that Lysine was first suggested by @Bdeep86.


Have a look at the following Google Sheet :


https://docs.google.com/spreadsheets/d/1CLtqxW0-L8f25ZXD2H6FFDVbggc73GGC8KtYNkjdUBo/edit?usp=sharing


The sheet lists the various Liver issues that can set the Stage for ME/CFS. It will be constantly updated.


Please also look at the following video where Dr Alan Light discusses about ME/CFS :

Dr Light talks about "an underlying susceptibility" and that "then something happens". In this case (=underlying susceptibility) we may have a Cholestatic event, Hemochromatosis,Viral Infection, Gallstones, Gilbert's Disease etc.

As an example, Hemochromatosis generates Oxidative Stress and impairs Liver function. Now consider an individual that has this susceptibility : The liver is able to compensate until something additional happens (EBV, Medication, Hepatitis, etc) that further disrupts liver function and after that the body falls into a vicious cycle of Oxidative Stress, Inflammation and Autoimmunity.

We now move to an other video by Dr. Alan Light :

So he speaks about HSP6 which is a Heat shock protein. Heat shock proteins are related with proper Protein folding and the Thread "Unfolded Protein Response" has many mentions on Heat Shock Proteins since 2015.

According to Genecards about HSP6:

Molecular chaperone implicated in a wide variety of cellular processes,including protection of the proteome from stress, folding and transport of newly synthesized polypeptides, activation of proteolysis of misfolded proteins and the formation and dissociation of protein complexes. Plays a pivotal role in the protein quality control system, ensuring the correct folding of proteins,the re-folding of misfolded proteins and controlling the targeting of proteins for subsequent degradation.


So to recap : Using Machine Learning we were able to identify the importance of Heat Shock proteins, Pyruvate Dehydrogenase Complex and impaired Phospholipids Metabolism well before any other ME/CFS Researcher.

Unfortunately, there are still many questions that have not been answered :

a) What is the prevalence of all of the Liver stressors that are listed to the Google sheet to ME/CFS Patients?
b) What is the prevalence of Liver Fibrosis to ME/CFS Patients (especially those with > 5 years having ME/CFS)

Despite extensive Testing of ME/CFS Patients, no Fibroscans or even Total Bile Acids (a very simple blood test) have been performed. I sent some fibroscans and Total Bile Acids tests in hope to convince Researchers to look more closely at the Liver/Gallbladder function.

I will keep trying.

 

[aquariusgirl]

@mariovitali @Bdeep86, Check out Steven Zaisel, at UNC>...... his theory about choline insufficiency syndrome.. may explain the way in for some of us.. and since estrogen regulates PEMT .. it would explain why so the X factor.. why women get this illness.

I wonder how many women on here.. have fatty liver syndrome? I'm guessing there's a subset of us who have fatty liver, choline deficiency, choline associated POTs etc

also, wondering if there is a role for metalloestrogens perturbing, disturbing estrogen metabolism, interfering with estrogen regulation of PEMT....which then impacts phosphatidylcholine production.....and hence mito function?

 

[NotThisGuy]

My POTS got significantly better when I started smoking tobacco again.
so yeah, there might be a choline connection

 

[aquariusgirl]

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4415014/

 

[mariovitali]

Here is the latest update :


-The software confirms the importance of Sepsis. This has been forwarded to @Janet Dafoe (Rose49) along with some other associations that the software has made regarding Sepsis.

Across all algorithmic runs, Sepsis comes to the second position after "Liver Disease" :

"mtdna" above referres to "Mitochondrial DNA".

Here is a snapshot of the associations with Sepsis that the tool has found: The tool suggests that Inflammatory response is mostly associated with Sepsis (a thing which we already know) however other -possibly interesting- associations were also found. I forwarded to Janet also the possible involvement of Vitamin K to red cell deformability issues and also the importance of Glycoproteins (which was also mentioned to previous emails to Ron) :

Next, we have a possible involvement of LXR to Autism. Recall that LXR was identified as a possible target for ME/CFS by Machine Learning .

Dr. Naviaux suggests that Autism and ME/CFS may be sharing the same mechanism(s) :

http://forums.phoenixrising.me/index.php?threads/how-similar-is-autism-to-me-cfs.51797/


Here is the study linking LXR to Autism :


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4859958/


I will also send an e-mail to Robert Phair in hope that some findings not previously mentioned will help him with his "Metabolic Trap" hypothesis.

I will update you all.

 

[NotThisGuy]

@mariovitali nice.
to help me understand:
the results gut,LPS and microbiome are not worth mentioning? Like could the sepsis come from leaky gut for example?
cannabinoid receptor also showed up. Is there any further information which receptor it is and if its activation or inhibition is related to cfs?

 

[mariovitali]

@NotThisGuy

I sent these exact snapshots to @Janet Dafoe (Rose49) so any additional usage of the algorithmic findings are now up to the OMF Researchers on how /if they will use them.

 

[mariovitali]

Has there been any update regarding this Research @Janet Dafoe (Rose49) ???


https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5365380/

Primary bile acids such as sulfoglycolithocholate (13) are synthesized in the liver and the major bile salts result from its conjugation with taurine (28) and glycine, forming taurochenodeoxycholate and glycochenodeoxycholate (21), respectively, which are all numbered metabolites found to be significantly reduced in ME/CFS patients. Reduction in these compounds is suggestive of damage to the liver. A study from the FDA National Center for Toxicological Research (NCTR) was able to identify liver injury biomarkers as the result of drug-induced hepatotoxicity in rats25. Strikingly, several other metabolites identical to our findings were also identified in their report, namely 5-guanidino-2-oxopentanoic acid (12, also named 2-oxoarginine), sebacic acid (60), along with energy metabolites from the glyoxylate and dicarboxylate metabolism. These biomarkers could be used to define a serum metabolic signature by creating a panel for hepatotoxicity prediction25.

and

The results of our pilot study on a small cohort is promising with regard to the possible use of plasma metabolite profiling for diagnosis of ME/CFS. After our manuscript was submitted for review, two other independent studies appeared with results that concur with our findings, while sampling distinctive populations. Indeed, Naviaux et al.32 found a similar trend of hypometabolic state in their study, with over 80% of significantly different metabolites being decreased in patients vs. controls. Although the compounds they measured are not all identical to the ones we detected, the major pathways determined to be affected by ME/CFS agree with our conclusions, and primarily include lipid metabolism and amino acids.

A summarized comparison with their female cohort is presented in Table 6. An additional study33with stronger statistical power, thanks to a larger cohort along with a validation dataset, also points to deficiencies we noticed, namely the urea cycle which takes place in the liver, and the TCA cycle which in part generates energy from lipids and sugars to make ATP, all compounds/pathways we found to be affected in our CFS cohort when compared to healthy controls. An even larger study with the latest methods for discovery metabolomics might reveal a larger set of metabolites that can distinguish cases from controls with greater accuracy, as well as allowing us to generate additional hypotheses regarding metabolic dysregulation in ME/CFS.

Apparently they use Classification methods (the same that were used for my Research) to identify important features.


Do we have any comment from Ron regarding this paper?

 

[mariovitali]

@Learner1

Regarding your question on how the Network Analysis has being generated, the idea was simple : Identify all Topics discussed in ME/CFS Research (such as T-Cells, NAD, Mitochondria, Liver, Vagus Nerve and many many more) that exists in PubMed using Natural Language Processing and use Network Analysis to see which topics are central to the problem at hand ( this is a simplistic way to explain it)

I just want to say again that the methods used have identified the following well before any other ME/CFS Researcher :

a) Pyruvate Dehydrogenase (Mella et al)
b) Bile Acids, Fatty Acids impaired Metabolism (Naviaux)
c) Potential Liver Involvement (Hanson)

Here is an older version of the Network :

You see some Nodes not visible. What i can disclose is that most of them are related to Liver function.


Now see this :

Interestingly, Network Analysis on the first figure contains the following nodes which also exist on the second figure : Oxysterols, PPAR, LXR

Given that Gas6, Mer and Pros (shown on the second figure) are Vitamin K related Genes, CYP27A1 is associated with Bile Acid Metabolism (and as a result Fatty acid metabolism), that CYP7B1 has to do with Cholesterol metabolism and that this setting is associated with Inflammation and Autoimmunity is an interesting coincidence.

 

[Lisa108]

@mariovitali, this is wayyyy over my head, but thank you for your efforts!

 

[mariovitali]

@Lisa108

Perhaps something more useful would be to see the following spreadsheet and make sure you haven't had any of these conditions (two tabs exist)


https://docs.google.com/spreadsheets/d/1CLtqxW0-L8f25ZXD2H6FFDVbggc73GGC8KtYNkjdUBo/edit?usp=sharing

 

[Learner1]

@Learner1

Regarding your question on how the Network Analysis has being generated, the idea was simple : Identify all Topics discussed in ME/CFS Research (such as T-Cells, NAD, Mitochondria, Liver, Vagus Nerve and many many more) that exists in PubMed using Natural Language Processing and use Network Analysis to see which topics are central to the problem at hand ( this is a simplistic way to explain it)

I just want to say again that the methods used have identified the following well before any other ME/CFS Researcher :

a) Pyruvate Dehydrogenase (Mella et al)
b) Bile Acids, Fatty Acids impaired Metabolism (Naviaux)
c) Potential Liver Involvement (Hanson)

Here is an older version of the Network :


View attachment 27064



You see some Nodes not visible. What i can disclose is that most of them are related to Liver function.


Now see this :

View attachment 27065
Interestingly, Network Analysis on the first figure contains the following nodes which also exist on the second figure : Oxysterols, PPAR, LXR

Given that Gas6, Mer and Pros (shown on the second figure) are Vitamin K related Genes, CYP27A1 is associated with Bile Acid Metabolism (and as a result Fatty acid metabolism), that CYP7B1 has to do with Cholesterol metabolism and that this setting is associated with Inflammation and Autoimmunity is an interesting coincidence.

@mariovitali I wasn't asking how the analysis was generated. I'm interested in getting well, so would like to understand how this wonderful data you're generating can be used to verify that that is indeed what's wrong with us, in all its complexity, and create treatment plans that fix us.

 

[mariovitali]

@Learner1


CC : @Murph @Lisa108 @Violeta @Jesse2233


I am doing anything i can for this. What i can tell you is that no one has dismissed it as a Theory and in fact they told me that this Theory is quite interesting. I won't go into names but they are all Professors in Medicine (amd i am not talking about Prof. Davis). If they choose to make me a part of their Team i will disclose who these people are.

The theory goes like this :

a) All ME/CFS Patients have less than optimal Liver and/or Gallbladder function.
b) A Liver Stressor (EBV, some Medications, Hepatitis and many more) disrupts key Pathways in Oxidative Stress, Endoplasmic Reticulum Stress , Autophagy, Bile Acid Metabolism and may be more.
c) The body falls into a vicious circle of Oxidative Stress, Inflammation, Autoimmunity and Unfolded Protein Response events.
d) The only way to get out of this vicious circle is personalised regimen

I will like now to present for the first time an e-mail i sent to a number of ME/CFS Researchers in 2015. This is not about pointing the finger to anyone. This is about saying that for the past 3 years i've been trying *really* hard to bring a solution. And i will not quit because i believe it may help:

-Interestingly i talk about "a vicious cycle" 3 years ago.
-I also talk about SNPs in Multiple Pathways being at play 3 years ago.
-I also said that the solution may be a Personalised Regimen at a later time (not shown here). I am willing to work with Researchers to prove this point.

Please add to this the fact that these techniques have found many important pieces of the puzzle before other Researchers.

See here for a list of Liver conditions that may be setting the stage for ME/CFS (and many more syndromes)

https://docs.google.com/spreadsheet...02512741855408673597&usp=sheets_home&ths=true

Rest assured, i will keep trying.

 

[Learner1]

I agree with your theory. It fits with what the metabolomics researchers have been finding. What do we do about it, though? We need effective and efficient treatments.

 

[mariovitali]

@Learner1

I understand what you are saying but the evaluation of this Theory needs to be done from Scientists. This personalised regimen needs to be performed to patients under controlled conditions and under Medical supervision. This is what i am trying for the past 3 years, to convince Researchers to accept and try out this Theory.

The regimen consists of the following :

a) Avoidance of certain foods and substances (that fit the individual). generally speaking P450 Inducers are ok, P450 Inhibitors are problematic
b) Supplementation with Vitamins, Herbs and Minerals (that fit the individual)
c) Constant Monitoring of the Symptoms to evaluate and change the regimen acordingly if necessary

 

[Learner1]

Oh, that's what my doctor and I are doing.

 

[mariovitali]

Oh, that's what my doctor and I are doing.

I hear what you are saying but there are different ways to do "personalised" regimen. I read in one of your posts that you take Curcumin. Believe it or not this may be not good for you in the long run, despite being an anti-inflammatory.

Apart from this, you must find if a "Liver stressor" is at play. Have you checked yourself for Hemochromatosis for example? If yes, unfortunately there is a huge list of tests that you should do in order to completely rule out all the different things that may be affecting your Condition.

 

[Diwi9]

@mariovitali - I have a vague understanding of how your report was generated. What are the limitations to this effort? For example, it is dependent on studies that have been conducted, so that emerging work (like looking at neuroinflammation) may under influence the results? Have you attempted your algorithm by excluding certain studies or have a priority-ranking system for other studies? I'm just curious if you have looked at different methodologies and what the limitations are to your current methodology. It's great that researchers are paying attention to your work and I hope you can link up to continue your work with some collaboration.

 

[mariovitali]

@Diwi9

Thank you, you place a number of points that are valid. To answer your questions :

-The latest studies are being collected , analyzed using Natural Language Processing and added so that they are considered algorithmically. The process takes close to 3 days to complete every time a new Medical Topic is being added (...)

-At present there is no priority-ranking system (so for example prioritizing latest vs older research) or Research performed by experts in their field vs new scientists and so forth. But this is a good idea.

The thing is that it appears that the system learns at an unprecedented rate for other diseases as well and has come up with some hypotheses (which in my opinion do hold) which i would really like to show and share with researchers.

However, the Software should prove its findings on ME/CFS first before moving on to other diseases and conditions.

 

[Learner1]

I hear what you are saying but there are different ways to do "personalised" regimen.

Doing metabolomics testing to understand one's individual biochemistry is essential. Having an understanding of one's genes, environmental exposures, etc. is critical, too. Then, a clever doctor can personalize a plan that meets the individual's needs.

Apart from this, you must find if a "Liver stressor" is at play. Have you checked yourself for Hemochromatosis for example? If yes, unfortunately there is a huge list of tests that you should do in order to completely rule out all the different things that may be affecting your Condition.

We are all unique individuals with unique genes and environmental factors. It is pretty impossible for scientists to come up with sports of identical patients to run experiments are. We are not identical widgets.

Yes, I have hemachromatosis, and it has been treated. I had some other things, like chlamydia pneumoniae and heavy metal toxicity stressing my liver.

You make a good point that genetic conditions should be identified in individuals. I also have a lot of other interesting genes that may be impacting my situation, as other patients do, too.

I read in one of your posts that you take Curcumin. Believe it or not this may be not good for you in the long run, despite being an anti-inflammatory

As for curcumin, I have benefited in multiple situations from it. It has different mechanisms of action, which I have found to be extremely effective, and which are based on sound science. It is a COX-2 inhibitor, is a broad spectrum Phase II detoxifier, and has cancer-fighting properties. I have seen it reverse cancers that have been resistant to multiple chemo regimens. In my case, it has helped with removing toxic arsenic metabolites stuck halfway through the detox process. It has also calmed down brain and intestinal inflammation caused by intravenous immunoglobulins.

I suspect your network needs a few more data points to take into account all of my issues before recommending whether treatments that will or won't work. Real life is more complex than most studies.