With a nudge from AI, ketamine emerges as a potential rare disease treatment

It feels like in the last few years I have been constantly hearing about new research showing “fun” drugs as cures to a wide array of medical conditions. There’s marijuana for chronic pain, mushrooms for depression, mdma for ptsd, and now ketamine. They are often talked about as if they have little or no harmful side effects. I’m glad we are moving away from the incredibly destructive drug policies that have been in place for so long but I fear the pendulum may be swinging too far the other way. The opioid epidemic was largely caused by the idea of them being non addictive being heavy pushed by the pharmaceuticals industry. Are we going to see similar harms from these drugs a few years after they gain more popularity? So far the evidence doesn’t show anything near this level of harm. I’m just skeptical of anything that seems too good to be true.

&gt; Proof-of-concept for reasoning over medical knowledge graphs, using miniKanren + heuristics + indexing.<p><a href="https:&#x2F;&#x2F;github.com&#x2F;webyrd&#x2F;mediKanren" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;webyrd&#x2F;mediKanren</a>

Some 15 years ago someone (sorry I forgot who) in Waterloo university published a paper about how to use a search engine to discover hidden relations.<p>It was something like this:<p>1. Search for the key word &quot;A&quot;<p>2. Find new keywords (b,c,..) that are associated with &quot;A&quot; (I did this with lingpipe [0]<p>3. Search each the new keywords (b,c, ...)<p>4. Find for each of them their associated keywords (finding (s, t, ...)<p>5. Repeat 3.<p>6. repeat 4. finding (z, x, ...)<p>7. If in (z, x, ...) one of them is associated with &quot;A&quot; then you discovered an hidden relation between &quot;A&quot; and one of the members of (s, t, ...) (hidden if it did not appear in the initial search).<p>[0] <a href="http:&#x2F;&#x2F;www.alias-i.com&#x2F;lingpipe&#x2F;" rel="nofollow">http:&#x2F;&#x2F;www.alias-i.com&#x2F;lingpipe&#x2F;</a>

I am surprised people here seem mostly interested in discussing drugs when the object of this article is the materialization of the wet dreams of all the ML fanboys out there. Reasoning over medical information that actually works! Isn&#x27;t that remarkable?! Of course, it&#x27;s symbolic and not statistical &#x27;AI&#x27; so it doesn&#x27;t tickle the deep learning hipsters announcing AGI as just around the corner.

&quot;The tool, called mediKanren, scanned millions of biomedical abstracts hunting for relationships between existing compounds and the gene involved in the disease.&quot;<p>Is this AI or someone put the effort into compiling a database and added automatic search function?

<i>&quot;In these patients, that gene is sort of running at 50% capacity: One copy’s broken, the other copy’s working,” explained Might. “So the tool, when it sees that, says, ‘OK, well, maybe I can make the functioning copy work twice as hard to compensate.’”</i><p>Woo! Biohacking at its best.<p>I do something sort of similar by eating hot peppers regularly. It opens up a different chanel than my defective chanel and helps me keep my chronic edema down to a dull roar, among other things.<p><i>To prepare the kids and their parents, the Seaver center sent stickers to practice applying to the kids’ chests, so they could get comfortable with the sensory experience of ECG readings.</i><p>I like these people.

I recently got my full genome sequenced via a website called nebula.org. I don&#x27;t have any backgrounds in genetics but I was playing around with the raw data with some github apps. Thanks so much for this - I may be able to link this app in to my custom data. That would be so awesome if we could hook this up to personal genetics results. Even people who aren&#x27;t sick do have mutations and it would be so cool seeing AI running some kind of search on their personal genome and suggesting a drug for that!<p>Here&#x27;s the link to the github app they mentioned: <a href="https:&#x2F;&#x2F;github.com&#x2F;webyrd&#x2F;mediKanren" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;webyrd&#x2F;mediKanren</a>

On reading this article, it must be so hard for parents to see a way forward for treating their child&#x27;s condition, but be roadblocked by either funding, or study participation criteria. If I observed my child have such a profound change as Mateo did, I think I might be incentivised to continue seeking these drugs for their treatment. I know that it would be absolutely a criminal act to purchase ketamine illegally and administer it to my child, or to find a shady doctor who prescribe it off-label. I wonder how often these scenarios come up? And how often parents go out of their way to seek unapproved medicines (much like the AIDS Buyer&#x27;s Clubs of the 90s).

<i>“I advise strongly against trying to pursue ketamine, commercially or clinically, because of safety concerns and the risk of losing our opportunity to study it rigorously,” said Kolevzon</i> [the clinical director]<p>Research to gain quality evidence is good, but sometimes other considerations outweigh that. Visible improvement in the sufferers of a rare disease may be such a case.

The video that explains it all for techies: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;d-Klzumjulo" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;d-Klzumjulo</a>