Interesting new data

[dl2dp2]

Advertisement - Members don't see this ad

Freshly saw this through NIH's mass mailing today:

NIH Extramural Nexus (News) | Grants & Funding

nexus.od.nih.gov

 

[SurfingDoctor]

First R01 20 years after matriculation from a T32?! Realizing that the inflection point is more at the 10 year mark, but still. That's some dedication, but really at that juncture, what's the point?

 

[mark-ER]

Eh, you never know. Some people go into private practice after T32, and then change their mind again. No one is judging.

 

[tr]

First R01 20 years after matriculation from a T32?! Realizing that the inflection point is more at the 10 year mark, but still. That's some dedication, but really at that juncture, what's the point?

This is only weird if you imagine a research career is synonymous with continuous R01 support. It's probable that some people in that 20-y group may have left science and later returned, but I would imagine many of them had relatively continuous research careers that happened to be supported by non-R01 type funding (e.g. CDC, PCORI, European mechanisms, industry/pharma, foundation, or even just non-R01 NIH mechanisms like R21, R34, U series etc).

Thinking about the biosketches I've seen lately, I don't see too many people who are continuously supported on back-to-back R01s. Most of them have a much broader mix of funding in their histories. I have a family member who is a basic scientist and has been on the same two R01s continuously renewed for the past 30+ years, but that's not a model I see too much in the more clinical/translational environment where I work. People in my area with R01s seem to mostly have them for a specific ~5-year project which is over when it's over.

 

[SurfingDoctor]

This is only weird if you imagine a research career is synonymous with continuous R01 support. It's probable that some people in that 20-y group may have left science and later returned, but I would imagine many of them had relatively continuous research careers that happened to be supported by non-R01 type funding (e.g. CDC, PCORI, European mechanisms, industry/pharma, foundation, or even just non-R01 NIH mechanisms like R21, R34, U series etc).

Thinking about the biosketches I've seen lately, I don't see too many people who are continuously supported on back-to-back R01s. Most of them have a much broader mix of funding in their histories. I have a family member who is a basic scientist and has been on the same two R01s continuously renewed for the past 30+ years, but that's not a model I see too much in the more clinical/translational environment where I work. People in my area with R01s seem to mostly have them for a specific ~5-year project which is over when it's over.

I suppose that possible. Most biomedical research is funded by the NIH, but certainly there is foundation support. Though I thought that was dwindling too? It is certainly more common than it used to be of having gaps in funding. Either way, it would certainly be interesting to know for those late first R01s, where the support came from.

 

[CheGuevaraMD]

I suppose that possible. Most biomedical research is funded by the NIH, but certainly there is foundation support. Though I thought that was dwindling too? It is certainly more common than it used to be of having gaps in funding. Either way, it would certainly be interesting to know for those late first R01s, where the support came from.

"Matriculation" means enter, so matriculating into a postdoctoral T32 is equivalent to graduating from graduate school or medical residency. Given that postdocs are 4-8 years long, and K awards are 4 years long, it's not that crazy that people take at least 10 years to win an R. Twenty is pretty crazy, but plenty of MD-PhDs maintain a loose connection to research throughout their career, and plenty of PhDs can get tenure at lower-ranked institutions that do not require an R.

 

[SurfingDoctor]

"Matriculation" means enter, so matriculating into a postdoctoral T32 is equivalent to graduating from graduate school or medical residency. Given that postdocs are 4-8 years long, and K awards are 4 years long, it's not that crazy that people take at least 10 years to win an R. Twenty is pretty crazy, but plenty of MD-PhDs maintain a loose connection to research throughout their career, and plenty of PhDs can get tenure at lower-ranked institutions that do not require an R.

I suppose that’s fair. I guess the people I know, including myself, had about 5 or 6 years from entering into a T32 and having a K, and another 5 years between starting a K and starting an R, give or take a year or two. My experience is also colored by the “up or out” by year 8 of faculty if there’s no R award, which clearly varies amongst institutions.

 

[SurfingDoctor]

There's probably a better thread for this, but I'm too lazy to search...

NIH Extramural Nexus (News) | Grants & Funding

nexus.od.nih.gov

I guess I would have thought after 25 years, the trends would have stabilized... but then again, that's why I don't gamble in Vegas.

(also, honorable mention to the people in their late 50s getting their first R01. I seriously really admire that persistence, truly)

 

[SurfingDoctor]

I guess I can put this here...

As referenced in the most recent OpenMike blog...

Inequalities in the distribution of National Institutes of Health research project grant funding

Funding inequalities for NIH-supported research project grant investigators and organizations have increased over the past 25 years, but have modestly decreased over the past 3 years.

elifesciences.org

Figure 2... oof

Also:

Characteristic​		Top 1%​	Bottom 99%​

Funding in $Million

Median (IQR)

4.8 (4.0 to 6.5)

0.4 (0.3 to 0.7)

I hope the NIH is getting their money's worth out of those in the top 1%.

 

[Fencer]

I am collaborating with several other investigators and have a preliminary article in peer-review about that group of investigators (super-PIs).

 

[SurfingDoctor]

I am collaborating with several other investigators and have a preliminary article in peer-review about that group of investigators (super-PIs).

Maybe you don't want to spoil it... but are they Nobel laureates? I would imagine Jennifer Doudna is one of them.

 

[dl2dp2]

Characteristic​		Top 1%​	Bottom 99%​

Funding in $Million

Median (IQR)

4.8 (4.0 to 6.5)

0.4 (0.3 to 0.7)

I hope the NIH is getting their money's worth out of those in the top 1%.

I'm interested in the Gini index of the distribution to something that has more intuition (i.e. income of US, etc).

The median funding level as a PI is about 0.4M per year, so if you are above that you'd be an above-average PI. This is actually very useful info. If you want to end up at a top university you should probably be above median at least... which more or less is true.

 

[SurfingDoctor]

I'm interested in the Gini index of the distribution to something that has more intuition (i.e. income of US, etc).

The median funding level as a PI is about 0.4M per year, so if you are above that you'd be an above-average PI. This is actually very useful info. If you want to end up at a top university you should probably be above median at least... which more or less is true.

See there’s a rub there. They based this on total cost (directs plus indirects). I don’t know if they accounted for that in the methods (I didn’t read it that close). But my RPG beats that median only because of the indirects where my institution charges 70 cents on the dollar.. Hell, I’ll take what I can get, but since indirect rates are wildly different between institutions, that would automatically lead to institution inequality.

 

[Fencer]

Jennifer Doudna - active funding:

RePORT 〉 RePORTER

reporter.nih.gov

 

[dl2dp2]

I'm interested in the Gini index of the distribution to something that has more intuition (i.e. income of US, etc).

The median funding level as a PI is about 0.4M per year, so if you are above that you'd be an above-average PI. This is actually very useful info. If you want to end up at a top university you should probably be above median at least... which more or less is true.

Oooo!!! Including indirects: I beat the median!!

 

[SurfingDoctor]

Jennifer Doudna - active funding:

RePORT 〉 RePORTER

reporter.nih.gov

Exactly. She helped discover CRISPR-Cas9 system and got to share the Nobel prize… so I can understand why she gets so much.

Incidently, I remember reading somewhere that most Nobel laureates get their prize based on discoveries they made early in their career, like late 30s to late 40s or so. Makes the idea of giving the most money to the most senior people seems like a fools errand in that regard.

Edit: I guess that data is old. The average age in mid to late 50s. Well, still makes the idea of giving lots of money to 70 year olds unlikely to be fruitful but I’d love to see that data of how success is actually measured besides the most forgotten papers in journal X.

 

[SurfingDoctor]

Oooo!!! Including indirects: I beat the median!!

Your university must have nicer toilets than all those lowly low indirect schools. What a perk!

 

[SurfingDoctor]

Exactly. She helped discover CRISPR-Cas9 system and got to share the Nobel prize… so I can understand why she gets so much.

Incidently, I remember reading somewhere that most Nobel laureates get their prize based on discoveries they made early in their career, like late 30s to late 40s or so. Makes the idea of giving the most money to the most senior people seems like a fools errand in that regard.

Edit: I guess that data is old. The average age in mid to late 50s. Well, still makes the idea of giving lots of money to 70 year olds unlikely to be fruitful but I’d love to see that data of how success is actually measured besides the most forgotten papers in journal X.

Just to follow up on this, the paper that Jennifer Doudua co-published in 2013 based on NIH-funded research in 2012.

High-throughput profiling of off-target DNA cleavage reveals RNA-programmed Cas9 nuclease specificity - PMC

The RNA-programmable Cas9 endonuclease cleaves double-stranded DNA at sites complementary to a 20-base-pair guide RNA. The Cas9 system has been used to modify genomes in multiple cells and organisms, demonstrating its potential as a facile ...

www.ncbi.nlm.nih.gov

This paper and discovery that netted her the Nobel prize... cost the NIH in the 2012 fiscal year... $189,700. That's it. Now that is a good return on investment. I guess its a diamond in the rough from the NIH perspective, but it makes you wonder what the annual fiscal cost was for the biggest discoveries funded by NIH research in comparison to how much money was boondoggled on large single or multi-PI grants that have netted close to nothing.

🤔

 

[tortuga87]

According to Table 3 in the Inequalities article, 651 MD/PhDs who are less than or equal to 45 years old have an R01 or similar grant.

Table 5 of the Long Term Trends paper says that age 45 is about 70th percentile for first R01 award.

So that means 651 * 1.3 = 846 MD PhDs of the early career cohort are predicted to eventually get an R01.

According to AMA, about 600 students graduate MD PhD programs per year. If the average MD/PhD takes 8 years, then 846/600/8 = 0.176. So these back of the envelope calculations suggest that only 17.6% of the recent MD/PhD graduates are predicted to earn an R01 in their career.

 

[Fencer]

There are other paths. The NIH data prior to 2012-13 only listed the contact PI. By age 45, I had earned a K08, VA Merit award, and was co-PI of a R01. I eventually got a VA program project, and other NIH grants as program director (T32 x 2, KL2, and a P30 core). By their definitions, I counted as a K08 failure because they were using R01 contact PI as their benchmark. Many of us also do Team Science and contribute in other ways. If you measure citations per dollar of independent research grants, I am really high up. You might have an issue with citations, but my manuscripts have changed FDA labels of drugs used by tens of millions and even recommended practice guidelines.

Here is an older thread from Open Mike:

• Open Mike 2016 - Citations per dollar (human vs non-human studies)

 

[SurfingDoctor]

Back to the dollar thing and the 1%. That prior paper showed $400000 as the median of the low 99%. This came across my NIGMS blog today...

The R35 is the mechanism here is the NIGMS version of the R35 that came out starting about 5 years ago, but the R01 data is what I'm talking about. In the pre-MIRA (pre-2016 or so, you really can't trust the post-data because the numbers are fudged by design) cohort, the median direct cost of an R01 was less than $200000, and even then has a pretty sharp tail. So when you are talking about a RPG with direct costs of >$400000, even those are huge statistical outliers. Granted, the earlier data was award per PI and this is just per award mechanism, but still.

 

[Neuronix]

What's the source on that chart? The R01 funding doesn't make a whole lot of sense to me. Max annual directs budget on R01 at NCI is $500k and it seems like most grants try to max that out. Is NIGMS R01 budget smaller?

This figure (Inequalities in the distribution of National Institutes of Health research project grant funding) makes more sense to me. Most R01 or similar grants in the $250-500k range with some people with multiple large grants or large project grants.

 

[SurfingDoctor]

What's the source on that chart? The R01 funding doesn't make a whole lot of sense to me. Max annual directs budget on R01 at NCI is $500k and it seems like most grants try to max that out. Is NIGMS R01 budget smaller?

This figure (Inequalities in the distribution of National Institutes of Health research project grant funding) makes more sense to me. Most R01 or similar grants in the $250-500k range with some people with multiple large grants or large project grants.

MIRA Renewals: Award Rates and Budget Changes

Biomedical Beat highlights NIGMS-funded research and scientists, as well as images and tools that help illustrate biomedical concepts and advances to the general public.

loop.nigms.nih.gov

I don't know about NIGMS directs compared to other institutes. They tend not to fund clinical trials, which can balloon the direct costs so maybe that's it? I don't really know.

In the graph you linked, as we were talking about above, those are based on total costs, not only directs.

From the paper: From the NIH IMPAC II database, we obtained PI-specific data on inflation-adjusted total-cost funding of Research Project Grants (RPGs), defined as those grants with activity codes of DP1, DP2, DP3, DP4, DP5, P01, PN1, PM1, R00, R01, R03, R15, R21, R22, R23, R29, R33, R34, R35, R36, R37, R61, R50, R55, R56, RC1, RC2, RC3, RC4, RF1, RL1, RL2, RL9, RM1, UA5, UC1, UC2, UC3, UC4, UC7, UF1, UG3, UH2, UH3, UH5, UM1, UM2, U01, U19, and U34.

 

[dl2dp2]

According to Table 3 in the Inequalities article, 651 MD/PhDs who are less than or equal to 45 years old have an R01 or similar grant.

Table 5 of the Long Term Trends paper says that age 45 is about 70th percentile for first R01 award.

So that means 651 * 1.3 = 846 MD PhDs of the early career cohort are predicted to eventually get an R01.

According to AMA, about 600 students graduate MD PhD programs per year. If the average MD/PhD takes 8 years, then 846/600/8 = 0.176. So these back of the envelope calculations suggest that only 17.6% of the recent MD/PhD graduates are predicted to earn an R01 in their career.

This is the correct number as I have numerously cited in various posts. It's between 15-20%. This number is also decreasing. High-end programs anecdotally it's maybe around 30%, though there's a steeper drop in recent cohorts in high-end programs, as several alternatives became more appealing (industry, high-end private practice/procedural specialties, etc).

However, try do this math for PhD programs in biomedical sciences or MD program graduates. The cited number is around 10% for PhDs who get into a T32 postdoc. This is MAYBE 30% of all PhDs. This number is 5%ish for PhDs. Odd ratio of PhD vs. MDPhD is like 0.3.

Conclusion: if you REALLY want to run a lab, statistically, and the data-driven advice, your best shot is to do an MD/PhD... which is REALLY counterintuitive, because generally, the rule of thumb is that if you only want to run a lab, don't do MDPhD. But the reality is that the best "track" to R01 is top MDPhD, top residency/fellowship, K, then R01. This track has an approximately 50% success rate from start to finish.

 

[MSTPadvocate]

To add to the discussion, the data is a bit old but I think still holds true. Only about 20% of investigators (all comers) hold 2 NIH grants and the numbers holding 3+ are very small.