NIH Funding 2024

Notes

proj |> ungroup() |>
  summarise(n_projects=n(),
            n_organizatoins=length(unique(org)),
            PIperProj_mean = mean(n_pi),
            across(c(direct=direct_cost_amt,
                     indirect=indirect_cost_amt,
                     awarded=award_amount),
                   c(total=\(x) sum(x,na.rm=T),
                     median=\(x) median(x, na.rm=T)))) |>
  fmt() |>
  t()

n_projects	79,374
n_organizatoins	2,888
PIperProj_mean	2.6
direct_total	29,847,146,202
direct_median	256,779
indirect_total	10,805,710,856
indirect_median	112,671
awarded_total	42,169,253,525
awarded_median	388,750

proj |> filter(letter %in% c('R','K','F')) |>
ggplot() +
  aes(x=direct_cost_amt, fill=letter) +
  geom_histogram(alpha=.7, position='dodge') +
  #geom_density(alpha=.7) +
  scale_x_continuous(trans='log10', limits=c(5000,NA)) +
  cowplot::theme_cowplot() +
  #facet_wrap(letter~.) +
  labs(fill="grant", x="direct cost (log)", title="distribrution of award by type")

Top

Institution

proj_org_smry <- proj |>
  group_by(org) |>
  summarise(
    amount=sum(award_amount),
    n_proj=n(),
    n_R=length(which(letter=="R")),
    n_K=length(which(letter=="K")),
    mean_n_pi = mean(n_pi),
    median_amount = median(award_amount),
    direct_cost=sum(direct_cost_amt,na.rm=T),
    indirect_cost=sum(indirect_cost_amt,na.rm=T))

proj_org_smry |> arrange(-amount) |> head() |>fmt()

	org	amount	n_proj	n_R	n_K	mean_n_pi	median_amount	direct_cost	indirect_cost
1	JOHNS HOPKINS UNIVERSITY	967,554,620	1,826	922	191	2.8	404,828	716,140,229	267,819,162
2	UNIVERSITY OF CALIFORNIA, SAN FRANCISCO	923,404,391	1,741	889	232	2.5	395,536	680,359,737	245,279,242
3	WASHINGTON UNIVERSITY	901,899,906	1,455	805	123	3.1	393,750	679,451,125	223,127,797
4	UNIVERSITY OF MICHIGAN AT ANN ARBOR	840,742,085	1,663	965	155	2.7	388,052	613,767,433	228,042,377
5	UNIVERSITY OF PENNSYLVANIA	790,934,580	1,557	824	132	2.6	406,043	560,715,831	235,792,132
6	UNIVERSITY OF PITTSBURGH AT PITTSBURGH	747,102,172	1,435	858	114	2.7	400,107	542,554,963	209,616,818

Contact PI

proj_pi_smry <- proj |>
  group_by(contact_pi) |>
  # remove first and middle name
  mutate(contact_pi=gsub(':.* ',':', toupper(contact_pi))) |>
  summarise(
    amount=sum(award_amount),
    direct=sum(direct_cost_amt,na.rm=T),
    n_proj=n(),
    mean_n_pi = mean(n_pi),
    median_amount = median(award_amount),
    org=substr(paste(collapse=";", gsub('(UNIVERSITY|OF|SCHOOL|INSTITUTE) ?','', unique(org))),0,100))

proj_pi_smry |> filter(amount>100) |>
  mutate(projects=cut(n_proj, breaks=c(0,1,2,5,10,50,Inf)),
         pis=cut(mean_n_pi, breaks=c(0,1,2,5,10,50))) |>
  ggplot() +
  aes(x=amount, fill=pis) +
  geom_histogram() +
  scale_x_continuous(trans='log10') +
  cowplot::theme_cowplot() +
  labs(fill="mean N co-pi", title="Amount per contact-PI")

summary(proj_pi_smry$amount)

     Min.   1st Qu.    Median      Mean   3rd Qu.      Max.      NA's
        1    211712    462500    851443    858892 341743406      1620

By total amount

proj_pi_smry |>
  arrange(-amount) |>
  head() |> fmt()

	contact_pi	amount	n_proj	mean_n_pi	median_amount	org
1	79478801:BRISCOE	341,743,406	43	1.0	1,117,108	LEIDOS BIOMEDICAL RESEARCH, INC.
2	10753426:NOLEN	289,804,000	2	1.5	144,902,000	RESEARCH TRIANGLE
3	10829359:GROSS	110,114,217	2	4.0	55,057,108	NEW YORK MEDICINE
4	78492086:MONTALVAN	66,600,000	2	1.0	33,300,000	WESTAT, INC.
5	1882258:BOXER	61,341,887	7	8.3	984,055	CALIFORNIA, SAN FRANCISCO;MAYO CLINIC ROCHESTER
6	6190835:DIAMOND	60,794,967	12	5.5	765,900	STANFORD ;MARYLAND BALTIMORE;WASHINGTON ;PITTSBURGH AT PITTSBURGH

By number of projects

proj_pi_smry |>
  filter(contact_pi!="NONE:") |>
  arrange(-n_proj) |>
  head() |>fmt()

	contact_pi	amount	n_proj	mean_n_pi	median_amount	org
1	79478801:BRISCOE	341,743,406	43	1	1,117,108	LEIDOS BIOMEDICAL RESEARCH, INC.
2	1891624:EBERLEIN	13,217,171	30	23	258,998	WASHINGTON
3	8497898:SHEPPARD	359,300	28	1	14,250	KEYSTONE SYMPOSIA
4	7039607:STEWART	2,027,360	25	1	40,000	COLD SPRING HARBOR LABORATORY
5	79112606:FREEDMAN	46,535,946	25	1	225,042	LEIDOS BIOMEDICAL RESEARCH, INC.
6	6774622:PASCHE	4,805,560	23	21	50,162	WAYNE STATE ;WAKE FOREST HEALTH SCIENCES

Code

See Makefile.

get_2024.py Uses the NIH reporter’s api, but goes by state (and DC + PR) to avoid return limits. This saves all output (>700Mb!) to a pickle file.

grants_to_csv.py parses the pickle to csv.