Brain imaging should probably include mention of neuroquant MRI techniques (though there have been improvements since neuroquant). Basically you want quantification of brain segment sizes on an absolute basis, this requires segmentation and can only be done noninvasively with MRI. MRI can also approximate diffusion which can point to trauma and to demyelination and lesions.
1 Like
Copied the section on neurologic markers to our document ā Longevity Biomarkers Draft - Google Docs
1 Like
Awesome, added some ideas.
Another useful search for context: https://elicit.org/search?q=strengths%2Fweaknesses+of+current+biomarkers+for+longevity. This tool also enables via NLP to ask questions about research and specific papers.
1 Like
Asked ChatGPT this question: What are current limitations of longevity biomarkers?
Longevity biomarkers are biological measures that are thought to be associated with an individualās lifespan or healthspan. These measures can be used to predict an individualās risk of age-related diseases and to track the effectiveness of interventions designed to extend healthy lifespan. However, there are several limitations to the use of longevity biomarkers.
One major limitation is that most longevity biomarkers are only weakly associated with lifespan. While some biomarkers, such as telomere length, have been shown to be predictive of lifespan in some studies, the relationship is often complex and not well understood. In addition, many longevity biomarkers are not specific to aging, and can be influenced by a wide range of factors, such as genetics, lifestyle, and environment. This makes it difficult to interpret the results of individual biomarker tests and to use them to make accurate predictions about an individualās lifespan.
Another limitation is that many longevity biomarkers are not easily measured in a clinical setting. Some biomarkers, such as telomere length, require specialized laboratory equipment and expertise to measure accurately. Others, such as biomarkers of metabolic health, require complex and invasive procedures, such as biopsies or blood tests, to obtain the necessary samples. This makes it difficult to use these biomarkers as routine tests in a clinical setting.
Finally, there is a lack of standardization and consistency in the use of longevity biomarkers. Different studies may use different biomarkers, or measure the same biomarker in different ways, which makes it difficult to compare the results of different studies and to draw conclusions about the effectiveness of different interventions. In addition, there is a lack of consensus on which biomarkers are the most reliable and predictive of lifespan, and on the best methods for measuring and interpreting these biomarkers.
Overall, while longevity biomarkers have the potential to provide valuable information about an individualās risk of age-related diseases and to track the effectiveness of interventions to extend healthy lifespan, there are still many challenges and limitations to their use. More research is needed to develop and validate reliable and predictive longevity biomarkers, and to develop more effective and practical methods for measuring these biomarkers in a clinical setting.
So cool!
2 Likes
yeah the future is using large language models and increasingly better scientific foundation models like https://galactica.org/ to rapidly build up ideas, ask scientific questions, get answers from different papers etc. and then to crowdsource enable smart community to participateā¦ Martin from https://www.lateral.io/ is building out a knowledge / discourse graph which we can explore for this topic and challenge
1 Like
Hello! Iām only halfway done with bloodwork panels. The topic appeared to be deeper than I thought.
Iāve gathered the most popular blood markers (+updated zotero) and now trying to somehow prioritise them. Iād be grateful if you could share your vision on this.
My current idea is to first take markers that highly correlate with chronological age and are present widely in publicly available datasets (this makes them useful for ML). Plus I want to take some complex/derived markers (e.g. Neutrophil-to-lymphocyte ratio) that are claimed to work good without ML.
1 Like
Hey! Thank you for the update. My suggestion would be to use your own judgment and write up a text describing things that you personally found interesting and useful. Since we agreed on 750 words/section, please keep it succinct and highlight key takeaway messages. Once we have all sections completed, weāll start optimizing.
2 Likes
Agree with @rpill
This is a review article, not a textbook (textbook may be phase 2, lol).
Goal is to provide VitaDAO community members (and others) a general grounding on longevity markers. In turn, this is expected to be the first step in standardizing markers for research.
1 Like
Hey! Copied the section on bloodwork markers (individual markers + several combined scores). Would someone give it a quick review please?
Longevity biomarkers review article + bounty : bloodwork panels
@evgenity Hi! Thank you for the update. The text largely overlaps with the section on laboratory biomarkers where we describe individual body fluid proteins/metabolites used to predict mortality/morbidity. The idea for this section would be to look into available biomarker combinations (not necessarily limited to blood panels).
The text is good, but you donāt say about panels much.
1 Like
I think I was a little confused by terms āpanelsā, letās sync to avoid misunderstanding. My current understanding:
- A āpanelā is a group of tests that are done to help diagnose a medical condition or to help identify potential health risks. (examples - Complete Blood Count panel, Basic metabolic panel, Comprehensive metabolic panel, Lipid panel)
- A āmodelā (or ML-model, scoring model) is an algorithm that uses combination of features to estimate potential health risks (examples - FRS, DOSI, QRISK3 and predictive ML-algorithms).
I hope we can merge my text to improve lab section - Iāve collected these markers from the articles that research marker combinations and their importance for ML-models. Iāll now continue to write a text about the models themselves (right?).
@evgenity By panel we mean a combination of parameters integrated to assess health status, predict risk, an outcome, etc. So in your terminology, a panel == a model. Yes, what we need for the paper is models.
We definitely will be able to use your text for the laboratory markers section.
3 Likes
Hey! Iāve copied the current version of ML-models section Combined scores, machine learning models
Iāll be grateful for ideas and additions)
2 Likes
Would be nice to compare the current tests from https://trudiagnostic.com/ to https://physioage.com/ and compare the literature which are most promisingā¦ more here: blueprint
2 Likes
The poor man (Bryan Johnson) is running on the wrong endpoints and eating drugs which are likely to shorten his life. Specifically, he takes lithium and is happy to see longer telomeres. But as we know telomeres are not supposed to get longer, and what he is looking at is an artifact due to having a different composition of blood cells. Meanwhile, the lithium is likely to contribute to a progressive loss of kidney function over time. Iām writing something on that. End of rant.
For the purposes of this paper, my argument is that if you want to measure telomere length over time, you need to define which cells you are measuring it in, instead of doing it in whole blood.
2 Likes
Fair point, think he is probably keen to collaborate on improving his measurements and approachā¦ one thought was throwing it into a discourse graphs and incentivizing literature review of the different interventions and his references of the scientific literature.
1 Like
A collaboration like that would be nice. (When I read my last message again it sounds snarky, I donāt mean to imply that Johnson isnāt doing his best to improve things.)
2 Likes
@bowtiedshrike @evgenity Hi! Been busy lately, but can finally allocate time to have this review finished. Letās do it!
2 Likes
Great to hear back from you! Thank you for pushing this forward. How do you see the next steps in the process? Ready to help.
1 Like
I will need to write the part on physiologic markers. Writing a section takes me about 15 hours; so if I write a few hours per day, itāll take me 1-1.5 weeks. I will post my updates here once Iām done.
3 Likes