Today’s roadmap

01
Preparing the data

Can we trust what goes into our analyses?

Codebooks · quality checks · SciDataReportR · Quarto
02
Measuring cognition

What does a cognitive score actually mean?

BRACE · processing · normative models
03
Learning from the data

What patterns can we discover?

Exploration · unsupervised learning · clustering
04
Building and validating models

Can we predict those patterns reliably?

Regression · random forests · validation · metrics
05
Explaining our models

What drives predictions, and what can we learn?

Variable importance · SHAP · DALEX

The purpose of computing is insight, not numbers.

Richard Hamming
Numerical Methods for Scientists and Engineers, 1962

01

Preparing the data

Can we trust what goes into our analyses?

  1. 01Preparing the data
  2. 02Measuring cognition
  3. 03Learning from the data
  4. 04Building and validating models
  5. 05Explaining our models

Clinical cohorts are powerful, but not analysis-ready

What makes them hard

  • The study was not designed around today’s question
  • Measures introduced at different visits; batteries change
  • Some variables collected only in a subgroup
  • Irregular intervals and selective follow-up
  • Each study protocol has its own structure and outcome definitions

Scientific validity: the research must be methodologically rigorous.

Emanuel, Wendler & Grady
JAMA, 2000

One of seven requirements for ethical clinical research.

Small decisions can dramatically change results

Merges A join silently drops 40 participants or duplicates 12 visits.
Missing data −9, −99999, and blanks may mean different things. Not collected, skipped, and refused are not equivalent.
Variable types A binary variable is treated as continuous.
Labels & definitions Cryptic export names appear in figures, or collaborators use different labels and definitions.
Statistical choices The statistical test doesn’t match the data or question.
Analysis history A result depends on filters or recoding decisions that were never documented.

None of these necessarily make the software throw an error. That’s what makes them dangerous.

To err is human; to really foul things up requires a computer.

Attributed to Paul R. Ehrlich

An analysis needs an inspectable history

Every transformation.
Every exclusion.
Every decision.

Someone else should be able to see what I did and reproduce it.

Visible, traceable, correctable.

What changed?

The recode, calculation, filter or model choice.

Why did it change?

The rule, rationale and relevant assumptions.

What did it affect?

The participants, measurements and results.

Replication is the ultimate standard by which scientific claims are judged.

Roger D. Peng
Reproducible Research in Computational Science, 2011

The data science pipeline is part of the science

SciDataReportR

Scan for the package and install instructions

Make it so.

Patrick Stewart as Jean-Luc Picard
Star Trek: The Next Generation

Why a package

  • The same steps, the same way, every project
  • Anyone in the group can see what we did and why
  • Change one decision, rerun everything
  • Open source: you can use it too
devtools::install_github("RDastgh1/SciDataReportR")
library(SciDataReportR)

github.com/RDastgh1/SciDataReportR

SciDataReportR: shared analysis and reporting foundation

Shared analysis and reporting foundation

MetabolomicsReportR hex sticker

In development

ProteomicsReportR hex sticker

In development

MultiPlexReportR hex sticker

In development

SciDataAgent hex sticker

SciDataAgent? Concept / in development

Shareable Codebook

Separating the code from the metadata

From codebook to labelled data, and a dictionary to share

out_Revalued <- RevalueData(data = df_Export, codebook = df_Codebook)   # apply labels, types and value codes
FormattedDataDictionary(out_Revalued$RevaluedData)                       # one row per variable, ready to send
Variable1,2 Label Type Ordered Factor n_missing complete_rate n_unique Top Counts Mean Median SD Min Max Histogram
HIVSTAT HIV serostatus factor FALSE 0 1 2 PWH: 1874, PWO: 1063
AGEATVIS Age (years) numeric 0 1 59 55.21 56 11.34 28 87 ▂▅▇▅▁
BSEDUCTN Education factor FALSE 10 1 5 Som: 942, Com: 629, Mor: 564, Les: 470
SEXBIRTH Sex at birth factor FALSE 0 1 2 Fem: 1552, Mal: 1385
EducationYears Education (estimated years) numeric 175 0.94 14 13.48 13 2.53 3 18 ▁▁▇▆▆
PHQ2 PHQ-2 depressive symptoms (0-6) numeric 31 0.99 8 1.22 0 1.59 0 6 ▇▂▁▁▁
1 Data Frame has 2937 rows and 6 columns
2 Color indicates type of variable

Failures in communication … resulted in a decision … based on incomplete and sometimes misleading information.

Presidential Commission on the Space Shuttle Challenger Accident
1986, Chapter V · Excerpt

Look before you model: what is missing, and why?

PlotMissingData(df_Sample, variables = vars_SampleMissing)

Above all else show the data.

Edward Tufte
The Visual Display of Quantitative Information, 1983

Longitudinal data: who was seen, and when?

PlotTimeSwimmer(df_Visits, id_var = "Participant", Time = "Date", State = "HIVSTAT", Format = "state_path", LineWidth = 0.65, PointSize = 1.7)

There is nothing better than a picture for making you think of questions you had forgotten to ask.

John W. Tukey & Paul A. Tukey
quoted in Friedman & Stuetzle, Annals of Statistics, 2002

A picture finds what a summary hides

out_Revalued <- RevalueData(df_SampleRaw, df_CodebookSample)
PlotContinuousDistributions(df_SampleAge,
  variables = c("AgeExport", "AgeRevalued"), ncol = 2,
  FacetLabelStyle = "label_only")

The greatest value of a picture is when it forces us to notice what we never expected to see.

John W. Tukey
Exploratory Data Analysis, 1977

Measure Mean Maximum
Age as exported 103.1 999
Age after codebook recoding 72.5 115

Explanation, code and results in one document

Allison Horst illustration: multiple programming languages feed a Quarto document, which produces HTML, PDF, Word and other outputs.

The code required to produce your output is part of the document itself.

Quarto Documentation
Tutorial: Computations

02

Measuring cognition

What does a cognitive score actually mean?

  1. 01Preparing the data
  2. 02Measuring cognition
  3. 03Learning from the data
  4. 04Building and validating models
  5. 05Explaining our models

BRACE: cognitive testing on a tablet

Flanker

Face task

BRACE in the MWCCS: Trail Making A and B, Stroop-Color, visuospatial learning. Animations: brainbaseline.com

What the tablet actually gives you

One file per task, for the whole study

Task phase Trial Stimulus colour Response colour Correct response Response time (ms)
Practice 1 Red Blue No 940
Practice 2 Blue Blue Yes 820
Test 1 Green Green Yes 760
Test 2 Red Red Yes 705
Test 3 Blue Green No 650
Test 4 Green Green Yes 810
Test 5 Red Red Yes 735
Test 6 Blue Blue Yes 790

There’s so much data out there, but there’s a finite amount of truth.

Nate Silver
WIRED, 2012

So we built an app for it

  github.com/BHPDataSci/BraceExtraction

Select the folder of raw BRACE exports
Documents
BRACE App › RawData_2025_03
Desktop
CancelOpen

A raw score is not a normative score

68 years old, did not finish high school Trail Making A: 22 seconds

T = 52

Illustrative people, scored with the published model.

The normative idea

T = 50 + 10 ×observed − expected for someone like youspread in the reference group
  • Expected comes from a reference group: here, people without HIV
  • Mean 50, standard deviation 10; higher is better on every test
  • A scoring algorithm is an interpretive model

A primary goal of clinical neuropsychological assessment is to determine whether a given set of test results suggests brain pathology.

Robert K. Heaton and colleagues
Demographic Influences on Neuropsychological Test Performance, 1996

Building a norm: one function

norm_TrailsA <- CreateNormativeTScoreModel(
  data            = df_BRACE,
  test_var        = "Trails_TrailsA_totalTime",    # the raw score: completion time in milliseconds
  covariates      = c("AGEATVIS", "Education"),    # what we adjust for
  reference_var   = "HIVStatus",                   # who defines "expected":
  reference_value = "PWOH",                        #   people without HIV
  convert_seconds = TRUE,                          # milliseconds -> seconds
  log_transform   = TRUE,                          # timed scores have a long tail of slow runs
  reverse_score   = TRUE                           # flip the sign, so higher always means better
)

norm_TrailsA$plots$raw

norm_TrailsA$plots$scaled

norm_TrailsA$plots$tscore

Applying the norm: what happens to one person’s score

df_Scored <- ApplyNormativeTScores(
  data          = df_NewPeople,
  normative_obj = norm_TrailsA
)

The saved model: coefficients from people without HIV

Term Estimate SE
(Intercept) -1.0661 0.0296
AGEATVIS -0.0041 0.0005
Education< High school -0.0256 0.0201
EducationSome college 0.0654 0.0170
EducationCollege 0.1264 0.0207
Education> College 0.1257 0.0181

Residual SD = 0.187. The outcome is −log10(seconds).

The same steps, applied to two people who both took 22 seconds

Person 1 · Raw (s) 2 · −log10 3 · Expected 4 · Difference ÷ SD 5 · T = 50 + 10z
32 y, college graduate 22 -1.342 -1.072 -1.45 36
68 y, < high school 22 -1.342 -1.372 0.16 52

The model is frozen The equation and every preprocessing step are stored in one object. The same object scores people with HIV, a later visit, another site or another study, so T-scores are comparable across papers.

Normative decisions change the apparent signal

Trail Making A. Grey line: no demographic adjustment. Orange: the norm you selected.

The most dangerous phrase in the language is, ‘We’ve always done it this way.’

Attributed to Grace Hopper

Published BRACE norms in the MWCCS

1,063

people without HIV: the reference group

MakeComparisonTable(df_BRACE, group_var = "HIVStatus", variables = c("AGEATVIS", "Education", "Sex", vars_T), AddEffectSize = TRUE)
Comparison by HIV serostatus (values: mean (SD)).
Characteristic PWOH
N = 1,0631
PWH
N = 1,8741
p-value2 Test2 Effect size ES method
Age (years) 56.95 (12.34) 54.22 (10.61) <0.001 Welch t-test 0.24 |d|
Education

<0.001 Pearson chi-squared 0.12 Cramer's V
    High school 201 (19%) 428 (23%)



    < High school 152 (14%) 318 (17%)



    Some college 307 (29%) 635 (34%)



    College 142 (13%) 180 (9.6%)



    > College 258 (24%) 306 (16%)



Sex at birth

<0.001 Pearson chi-squared 0.09 Cramer's V
    Female 499 (47%) 1,053 (56%)



    Male 564 (53%) 821 (44%)



Trail Making A T-score 50.00 (9.98) 49.98 (9.83) 0.954 Welch t-test 0.00 |d|
Trail Making B T-score 50.00 (9.98) 49.52 (10.06) 0.213 Welch t-test 0.05 |d|
Stroop-Color response time T-score 50.00 (9.98) 49.08 (10.14) 0.0177 Welch t-test 0.09 |d|
Stroop-Color accuracy T-score 50.00 (9.98) 48.96 (10.86) 0.00862 Welch t-test 0.10 |d|
Visuospatial learning (4x4) T-score 50.00 (9.98) 49.43 (9.88) 0.137 Welch t-test 0.06 |d|
1 Mean (SD); n (%)
2 p-values use Welch t-test (two groups) or ANOVA (more than two groups) for continuous outcomes and chi-square or Fisher’s exact test, selected from expected counts for categorical outcomes.

03

Learning from the data

What patterns can we discover?

  1. 01Preparing the data
  2. 02Measuring cognition
  3. 03Learning from the data
  4. 04Building and validating models
  5. 05Explaining our models

Exploring associations with cognitive T-scores

See, and then reason and compare and control. But see first.

William Osler

PlotSplitViolin(df_BRACEExplore, Var = "T_TrailsB", group_var = "POTUSE")

PlotAssociations(df_BRACEExplore, "POTUSE", "T_TrailsB")

PlotMiningMatrix(df_BRACEExplore, outcome_vars = vars_T,
  predictor_vars = c("CESD", "BMI", "PSS10", "NDRNKWK", "AGEATVIS", vars_BRACECategorical))
PlotAssociations(df_BRACEExplore, "CESD", "T_VSLT4x4")

PlotAssociations(df_BRACEExplore, "BMI", "T_TrailsB")

PlotCorrelationsHeatmap(df_BRACEExplore,
  predictor_vars = vars_BRACEContinuous, outcome_vars = vars_T)
PlotSplitCorrelationHeatmap(df_BRACEExplore,
  predictor_vars = vars_BRACESplitContinuous, outcome_vars = vars_T, split_var = "Sex")
out_PCA <- CreatePCAObject(df_BRACE, VarsToReduce = vars_T)

A global score can hide different cognitive profiles

I do not think it means what you think it means.

Inigo Montoya
The Princess Bride

What do we mean by impairment?

Why look at profiles?

Learning tasks and model choices

Choose the task before choosing the model.

Finding structure without labels: a two-level idea

Cluster the map, not the people

Cheaper. Cluster a hundred nodes instead of thousands of people.

Shape, not just level. Clustering raw test scores tends to find “high, medium, low”, which is the global score again.

The clustering pipeline: map, cluster, map back

Mclust groups the nodes into k clusters and gives every node a probability of belonging to each.

Map back gives each participant the cluster of her closest node.

How a map learns, one step at a time

Left: the data. Right: the same nodes on the grid. Toy data.

Watch the whole map learn

Grey: people · orange: nodes · lines join grid neighbours

Early: a wide neighbourhood moves the whole net. Late: only the winner moves.

Cognitive profiles in virally suppressed women with HIV

17

test T-scores, not domain scores

The design

  1. Self-organizing map of the test scores
  2. Mixture-model clustering of the nodes
  3. Random forests: what distinguishes each profile from unimpaired?

The map on real data: 929 women, 17 tests, 100 nodes

model <- CreateClusterModel_SOM_MClust(
  data = df_WIHS,
  variables = vars_NP,
  method = "exploratory",
  k_range = 2:10,
  som_xdim = 10,
  som_ydim = 10
)

Exploratory first Exploratory mode compares solutions. Once one is chosen, the same function refits it in "finalize" mode to give a model that can be applied to new people.

Cognitive Profiles

PlotClusterBoxplot(data = df_WIHS, ClusterVar = "Cluster",
  variables = vars_NP, ScoreType = "t", ReferenceLines = "t", ClusterLabel = "n")

PlotSpiderChart(df_WIHS, variables = vars_NP, group_var = "Cluster",
  ContinuousScaling = "none", interactive = TRUE) %>%
  plotly::layout(width = 1480, height = 610)

04

Building and validating models

Can we predict those patterns reliably?

  1. 01Preparing the data
  2. 02Measuring cognition
  3. 03Learning from the data
  4. 04Building and validating models
  5. 05Explaining our models

What is different between the groups?

MakeComparisonTable(df_WIHS, group_var = "Cluster", variables = vars_Key,
  Referent = "Unimpaired", AddPairwise = TRUE)
Comparison by Cluster (values: mean (SD)). Pairwise p-values are bonferroni-adjusted within outcome.
Characteristic Unimpaired
N = 3111
Executive function (sequencing)
N = 1291
Speed
N = 1441
Learning
N = 1371
Learning & memory
N = 861
Global weakness (processing speed)
N = 1221
p-value2 Test2 Unimpaired - Executive function (sequencing) Unimpaired - Speed Unimpaired - Learning Unimpaired - Learning & memory Unimpaired - Global weakness (processing speed)
Age 46.70 (8.07) 45.84 (8.65) 47.34 (8.94) 46.78 (8.46) 47.36 (8.08) 44.65 (9.26) 0.104 ANOVA 1 1 1 1 0.165
Education 12.80 (3.26) 13.02 (2.90) 12.47 (3.02) 12.39 (2.69) 12.13 (2.88) 12.87 (2.91) 0.192 ANOVA 1 1 0.83 0.316 1
Race





0.0451 Fisher (sim.) 0.0185 0.384 1 0.62 0.301
    black non-hispanic 202 (65%) 103 (80%) 104 (72%) 94 (69%) 65 (76%) 89 (73%)






    hispanic 45 (14%) 7 (5.4%) 15 (10%) 18 (13%) 12 (14%) 16 (13%)






    other 8 (2.6%) 5 (3.9%) 8 (5.6%) 3 (2.2%) 2 (2.3%) 6 (4.9%)






    white 56 (18%) 14 (11%) 17 (12%) 22 (16%) 7 (8.1%) 11 (9.0%)






Income





<0.001 Pearson chi-squared 0.726 0.0136 0.859 <0.001 0.121
    Less than 12k/y 126 (41%) 62 (48%) 80 (56%) 65 (47%) 60 (70%) 64 (52%)






    More than 12ky 185 (59%) 67 (52%) 64 (44%) 72 (53%) 26 (30%) 58 (48%)






Employed 152 (49%) 47 (36%) 41 (28%) 58 (42%) 23 (27%) 43 (35%) <0.001 Pearson chi-squared 0.085 <0.001 1 0.00127 0.0517
Depressive Symptoms





<0.001 Pearson chi-squared 0.218 0.00636 0.027 <0.001 0.0124
    Currently Depressed 66 (21%) 39 (30%) 51 (35%) 46 (34%) 39 (45%) 43 (35%)






    Not Depressed 245 (79%) 90 (70%) 93 (65%) 91 (66%) 47 (55%) 79 (65%)






Perceived Stress





0.00113 Pearson chi-squared 0.0151 0.286 0.934 <0.001 0.127
    Elevated Percieved Stress 230 (74%) 77 (60%) 94 (65%) 93 (68%) 44 (51%) 77 (63%)






    Not Elevated Percieved Stress 81 (26%) 52 (40%) 50 (35%) 44 (32%) 42 (49%) 45 (37%)






Post-traumatic stress 39 (13%) 32 (25%) 24 (17%) 27 (20%) 26 (31%) 19 (16%) <0.001 Pearson chi-squared 0.00765 1 0.215 <0.001 1
Crack, cocaine, heroin





<0.001 Fisher (sim.) 0.34 1 1 1 <0.001
    Current User 16 (5.1%) 4 (3.1%) 9 (6.3%) 8 (5.8%) 6 (7.0%) 5 (4.1%)






    Former User 170 (55%) 58 (45%) 79 (55%) 66 (48%) 46 (53%) 35 (29%)






    Never Used 125 (40%) 67 (52%) 56 (39%) 63 (46%) 34 (40%) 82 (67%)






Marijuana Use





0.00211 Pearson chi-squared 1 1 1 1 <0.001
    Former 170 (55%) 63 (49%) 77 (53%) 68 (50%) 49 (57%) 54 (44%)






    Never 78 (25%) 43 (33%) 45 (31%) 44 (32%) 24 (28%) 59 (48%)






    Recent 63 (20%) 23 (18%) 22 (15%) 25 (18%) 13 (15%) 9 (7.4%)






Nadir CD4 305.68 (235.61) 309.68 (237.58) 306.36 (229.22) 310.75 (217.87) 321.80 (214.21) 303.67 (205.78) 0.994 ANOVA 1 1 1 1 1
Current CD4 677.05 (343.19) 660.29 (316.82) 704.62 (312.29) 681.88 (310.92) 629.50 (286.48) 664.42 (288.29) 0.627 ANOVA 1 1 1 0.973 1
Years of effective ART 7.93 (4.75) 7.59 (4.59) 8.33 (4.29) 7.77 (4.59) 8.02 (4.49) 8.47 (4.45) 0.619 ANOVA 1 1 1 1 1
1 Mean (SD); n (%)
2 p-values use Welch t-test (two groups) or ANOVA (more than two groups) for continuous outcomes and chi-square or Fisher’s exact test, selected from expected counts for categorical outcomes. Pairwise p-values are bonferroni-adjusted within outcome.
Plot2GroupStats(df_WIHS, variables = vars_Model, impClust = vec_Profiles[1], normalClust = "Unimpaired", group_var = "Cluster")
Plot2GroupStats(df_WIHS, variables = vars_Model, impClust = vec_Profiles[2], normalClust = "Unimpaired", group_var = "Cluster")
Plot2GroupStats(df_WIHS, variables = vars_Model, impClust = vec_Profiles[3], normalClust = "Unimpaired", group_var = "Cluster")
Plot2GroupStats(df_WIHS, variables = vars_Model, impClust = vec_Profiles[4], normalClust = "Unimpaired", group_var = "Cluster")
Plot2GroupStats(df_WIHS, variables = vars_Model, impClust = vec_Profiles[5], normalClust = "Unimpaired", group_var = "Cluster")

Nonlinear relationships and interactions

Nowhere is it written on a stone tablet what kind of model should be used to solve problems involving data.

Leo Breiman
Statistical Modeling: The Two Cultures, 2001

Macchine learning models can pick up on nonlinear patterns

Start with one decision tree

From one tree to a random forest

Ways to validate a model

Trust, but verify.

Ronald Reagan
Quoting a Russian proverb, 1987

What will generalize?

Model selection, internal validation and testing in a new population answer different questions.

One partition to develop the model. One untouched partition for its final evaluation.

Each tile is one person. Splits must respect repeated visits, families or sites when relevant.

All participants one row per person Training set · 70% Test set · 30% Held aside during model selection Internal validation: fold 1 of 10 Fit on nine folds; assess on the remaining fold k-fold: rotate. Repeated k-fold: reshuffle and repeat. Repeat the fold assignment to assess variability. Fit the selected workflow using all training data Final held-out evaluation

Sample people with replacement. Fit on the resample and assess optimism or use observations left out of that draw.

Numbers show how many times each person was drawn. Zero means left out. Use a defined bootstrap validation estimator.

Outer resampling estimates performance of the complete selection process.

Inner resampling selects model settings using only the current outer training sample.

Outer training participants
Inner folds select the model settings
Refit on the outer training participants
Assess once on the outer assessment participants

Repeat across outer folds. Keep every model-selection decision inside the outer training sample.

Development cohort

Fit and select the workflow.

Freeze the scoring rules, transformations, model and decision threshold.

New setting

Evaluate in later participants, another site or an independent cohort.

Check calibration, discrimination and clinically relevant errors.

Internal resampling cannot establish transportability to a population it never observed.

A single performance number is incomplete

Does that encapsulate you?

Robin Williams as Sean Maguire
Good Will Hunting, 1997

Observed
Predicted In the profile Unimpaired
In the profile TP
True positive
FP
False positive
Unimpaired FN
False negative
TN
True negative
Sensitivity / recall=TPTP + FNHow many profile members did we find?
Specificity=TNTN + FPHow many unimpaired women did we identify?
Precision=TPTP + FPHow many positive predictions were correct?
Accuracy=TP + TNTP + TN + FP + FNHow many predictions were correct overall?
F1 score=2TP2TP + FP + FNHow do precision and recall balance?

The clinical question determines which errors matter. Performance also depends on the threshold and the population.

05

Explaining our models

What drives predictions, and what can we learn?

  1. 01Preparing the data
  2. 02Measuring cognition
  3. 03Learning from the data
  4. 04Building and validating models
  5. 05Explaining our models

Different profiles rely on different predictors

2026 reconstruction

SHAP: contributions to individual predictions

One example: the processing-speed profile versus the unimpaired group. Each point is an original participant.

Left: contribution away from the profile. Right: toward it. Categorical colors identify levels, without implying a biological order.

DALEX: cohort and individual response profiles

Across the cohort: which pattern did the model learn?

Centered response, in percentage points (ALE)

For one woman: how does her prediction change?

Predicted probability, with one example highlighted (ICE)

Left: a centered summary across the original cohort. Right: individual predictions with other inputs fixed. These describe the model, not treatment effects.

Guardrails: how not to fool yourself

Clusters are a description The algorithm will always return groups. Ask whether they are stable, and whether they predict anything you did not cluster on.

Performance before interpretation Importance and SHAP can be unstable when prediction is weak. Report validation and check rank stability.

Prediction is not causation A variable can be important because it stands in for something unmeasured.

Three things to take with you

1

The pipeline is part of the science. Merges, codes, quality rules and norms are scientific decisions. Make them visible and rerunnable.

2

A score is an interpretation. The reference group and the covariates decide who falls below the line.

3

Every summary compresses. Before collapsing to one label, ask whether the answer was in the variation.

4

Machine Learning Models are Hypothesis-generating. Not diagnostic.

Thank you

Leah Rubin, PhD
Jennifer Coughlin, MD
Yanxun Xu, PhD
Rebecca Veenhuis, PhD
Ana Soule, CRNP
Hannah Moore
Alison Buchholz, PhD
Kate Fitzgerald, ScD
Eran Shorer, MD
Jonathan Ye
Avery Ye
Wenxuan Lu
Lang Lang
K Ridgeway, PhD
Wei Jin, PhD
Cesar Tejera, PhD
Aparna Bhattacharyya, PhD

JH-BHP

Participants
and staff
Across Johns Hopkins
Collaborators
  • Erin Sundermann, PhD
  • David Moore, PhD
  • Scott Letendre, MD
  • Ron Ellis, MD
  • Qing Ma, PharmD
  • Rob Paul, PhD
  • Beau Ances, MD
  • Audrey French, MD
  • Kathleen Weber
  • Mary Clare Masters, MD
  • Luis Parra-Rodriguez, MD
  • Jane O’Halloran, MD, PhD
  • Shibani Mukerji, MD
  • Nicole Beaulieu Perez, PhD, RN
  • Helen Burgess, PhD
  • David Vance, PhD
  • Deanna Saylor, MD
  • Hemil Gonzalez, MD
  • Payal Patel, MD
  • Carrie Johnston, MD
Rebecca Veenhuis, PhD
Barb Slusher, PhD
Janice Clements, PhD
Amanda Brown, PhD
Ahmet Hoke, PhD
Alba Azola, MD
Rana Rais, PhD
Takashi Tsukamoto, PhD
Funding

R01MH125300 · R21MH125278 · MH113512 · NSF19-500 · MH119947 · MH120693 · P30MH075673 · P30AI094189 · R03MH123290 (Dastgheyb) · CFAR Scholar Award (Dastgheyb) · Michael and Ann Hankin Fund (Dastgheyb)

References