youainti
/
ClinicalTrialsPaper
1
0
Fork 0
Code Issues 6 Pull Requests Packages Projects Releases 1 Wiki Activity

Saving updates, added new causal graph

Browse Source
main
will king 3 years ago
parent 0e3e8e6b66
commit 8560435ec3
2 changed files with 152 additions and 71 deletions
Show Stats Download Patch File Download Diff File

186
Latex/Presentation/presentation.tex
Unescape Escape View File

@ -65,7 +65,7 @@ Washington State University \\ % Your institution for the title page
\vspace{12pt}
This research investigates the ways by which market conditions
My research investigates the ways by which competing compounds
affect clinical trial completion.
\end{frame}
%--------------------------------
@ -140,7 +140,7 @@ Washington State University \\ % Your institution for the title page
\frametitle{This research, in context}
In contrast to previous work looking at multiple phases of trials,
I seek to figure out what causes individual trials to fail.
I seek to figure out what causes individual (phase 3) trials to fail.
% \vspace{12pt}
%
@ -199,7 +199,7 @@ Washington State University \\ % Your institution for the title page
\end{enumerate}
\end{column}
\begin{column}{0.5\textwidth}
What does an \textbf{Terminated} trial look like?
What does a \textbf{Terminated} trial look like?
\begin{enumerate}
\item Study sponsor comes up with design
\item Apply for NCT ID from ClinicalTrials.gov
@ -381,8 +381,7 @@ Washington State University \\ % Your institution for the title page
\begin{enumerate}
\item Population Effects
\item Fundamental Safety and Efficacy of compound/dosage/route
\item How long the trial is taking
\item the total effects
\item How long the trial is taking.
\end{enumerate}
\end{frame}
%-------------------------------
@ -426,6 +425,20 @@ Washington State University \\ % Your institution for the title page
\cite{pearl_causality_2000}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Complete graph - Joint}
%introduce backdoor criterion
\begin{figure}
\scalebox{0.6}{
\tikzfig{../assets/tikzit/CausalGraph_joint}
}
\label{FIG:CausalGraphJoint}
\caption{Causal Graph with Joint RV}
\end{figure}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Sufficent Adjustment Set}
%introduce backdoor criterion
@ -438,7 +451,6 @@ Washington State University \\ % Your institution for the title page
\item Condition
\item Population
\end{itemize}
Discuss Regime Switching
For the direct effect these are controls for:
\begin{itemize}
@ -447,32 +459,30 @@ Washington State University \\ % Your institution for the title page
\item Population (optional)
\item Enrollment
\end{itemize}
Not causally identified due to Regime Switching
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Other testable hypotheses}
One advantage of this approach is that there are tools that can automatically
\begin{itemize}
\item verify causal identification
\item generate hypotheses to verify model
\end{itemize}
In this case, the following are testable hyptheses
\begin{itemize}
\item Decision to continue Phase III $\perp$ Elapsed Duration
\item Decision to continue Phase III $\perp$ Market Conditions $\Vert$ Condition
\item Decision to continue Phase III $\perp$ Population $\Vert$ Condition
\item Elapsed Duration $\perp$ Market Conditions
\item Elapsed Duration $\perp$ Condition
\item Elapsed Duration $\perp$ Population
\item Population $\perp$ Terminated $\Vert$ Condition, Decision to continue Phase III, Elapsed Duration, Enrollment Status, Market Conditions
\end{itemize}
\href{http://dagitty.net/mLyFuc5}{Dagitty.net model}
\end{frame}
%\begin{frame}
% \frametitle{Other testable hypotheses}
% One advantage of this approach is that there are tools that can automatically
% \begin{itemize}
% \item verify causal identification
% \item generate hypotheses to verify model
% \end{itemize}
%
% In this case, the following are testable hyptheses
% \begin{itemize}
% \item Decision to continue Phase III $\perp$ Elapsed Duration
% \item Decision to continue Phase III $\perp$ Market Conditions $\Vert$ Condition
% \item Decision to continue Phase III $\perp$ Population $\Vert$ Condition
% \item Elapsed Duration $\perp$ Market Conditions
% \item Elapsed Duration $\perp$ Condition
% \item Elapsed Duration $\perp$ Population
% \item Population $\perp$ Terminated $\Vert$ Condition, Decision to continue Phase III, Elapsed Duration, Enrollment Status, Market Conditions
% \end{itemize}
%
% \href{http://dagitty.net/mLyFuc5}{Dagitty.net model}
%\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Questions?}
@ -524,7 +534,8 @@ Washington State University \\ % Your institution for the title page
\begin{itemize}
\item USP Drug Classification (2023)
\begin{itemize}
\item Used to measure which drugs
\item A list of drugs that have certain effects.
\item Required manual matching between trials and categories.
\end{itemize}
\item NSDE Files (New drug code Structured product labels Data Element)
\begin{itemize}
@ -532,9 +543,8 @@ Washington State University \\ % Your institution for the title page
\end{itemize}
\item RxNorm
\begin{itemize}
\item Links pharmaceuticals between MeSH standardized terms and
NSDE files.
\item Used to find brand names that share active ingredients with those from trial.
\item Links pharmaceuticals between MeSH standardized terms,
NSDE files, and USP Files.
\end{itemize}
\end{itemize}
\end{frame}
@ -545,7 +555,6 @@ Washington State University \\ % Your institution for the title page
\item Global Disease Burden Survey (2019)
\begin{itemize}
\item Estimates of DALYs for categories of disease
\item Links of Categories to ICD-10 Codes
\end{itemize}
\item ICD-10 (2019)
\begin{itemize}
@ -556,14 +565,13 @@ Washington State University \\ % Your institution for the title page
\item Unified Medical Language System Thesaurus
\begin{itemize}
\item Used to link MeSH standardized terms and ICD-10 conditions
\item Manual matching process
\end{itemize}
\end{itemize}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Linking data}
%
The following linking process was used:
\begin{enumerate}
\item AACT trials to snapshots (internal ID)
@ -627,7 +635,7 @@ Washington State University \\ % Your institution for the title page
Note the implicit conditioning on trials treating diseases with IHME data\footnote{
IHME does not track data for W61.62XD: Struck by duck, subsequent encounter
}.
} or easily matched to codes and categories.
\end{frame}
%-------------------------------
@ -712,17 +720,26 @@ Washington State University \\ % Your institution for the title page
\item Allows for transfer of probability between groups
\end{itemize}
\end{itemize}
\begin{align}
y_i &\sim \text{Bernoulli}(p_i) \\
p_i &= \text{logistic}(X_i \vec\beta_{c(i)}) \\
\vec\beta_{c(i)} &\sim \text{MvNormal}(\vec\mu,\vec\sigma I)\\
\mu_c &\sim \text{Normal}(0, 0.05) \\
\sigma_c &\sim \text{Gamma}(4,20)
\end{align}
Gamma is using the shape and rate parameterization
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Total Effects Model}
\begin{align}
y_i &\sim \text{Bernoulli}(p_i) \\
p_i &= \text{logistic}(X_i \vec\beta_{c(i)}) \\
\vec\beta_{c(i)} &\sim \text{MvNormal}(\vec\mu,\vec\sigma I)
\end{align}
%TODO: describe X\beta
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Direct Effects Model}
%TODO: describe X\beta
\end{frame}
@ -748,6 +765,33 @@ Washington State University \\ % Your institution for the title page
%%%%%%%%%%%%%%%%%%%% Results
\subsection{Results}
%--------------------------------
%-------------------------------
\begin{frame}
\frametitle{Priors - Mu}
\begin{figure}
% \includegraphics[height=0.8\textheight]{../assets/img/Images/}
\label{FIG:Prior:Mu}
\caption{Prior - Mu}
\end{figure}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Priors - Sigma}
\begin{figure}
% \includegraphics[height=0.8\textheight]{../assets/img/Images/}
\label{FIG:Prior:Sigma}
\caption{Prior - Sigma}
\end{figure}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Priors - $p$}
\begin{figure}
% \includegraphics[height=0.8\textheight]{../assets/img/Images/}
\label{FIG:Prior:p}
\caption{Prior - $p$}
\end{figure}
\end{frame}
%--------------------------------
\subsubsection{Total Effect}
% - Review Parameter Values
@ -781,35 +825,35 @@ Washington State University \\ % Your institution for the title page
\frametitle{Questions?}
\end{frame}
%%-------------------------------
%%--------------------------------
%%\subsubsection{Direct Effects}
%%% - Review Parameter Values
%%% - hyperparameters
%%% - Table of MLE
%%% - Distributions
%%% - betas
%%% - Table of MLE
%%% - Distributions
%%% - Review Posterior Prediction for interventions
%%%--------------------------------
%%%-------------------------------
%%\begin{frame}
%% \frametitle{Convergence}
%% Sampling details
%% %TODO: UPDATE
%% \begin{itemize}
%% \item 6 chains
%% \item 2,500 warm-up, 2,500 sampling runs
%% \item seed = 11021585
%% \end{itemize}
%%\end{frame}
%%%-------------------------------
%%\begin{frame}
%% \frametitle{Questions?}
%%
%%\end{frame}
%%-------------------------------
%-------------------------------
%--------------------------------
\subsubsection{Direct Effects}
% - Review Parameter Values
% - hyperparameters
% - Table of MLE
% - Distributions
% - betas
% - Table of MLE
% - Distributions
% - Review Posterior Prediction for interventions
%--------------------------------
%-------------------------------
\begin{frame}
\frametitle{Convergence}
Sampling details
%TODO: UPDATE
\begin{itemize}
\item 6 chains
\item 2,500 warm-up, 2,500 sampling runs
\item seed = 11021585
\end{itemize}
\end{frame}
%-------------------------------
\begin{frame}
\frametitle{Questions?}
\end{frame}
%-------------------------------
%-------------------------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%% Conclusion %%%%%%%%%%%%%%%%%%%%%%%%
\section{Conclusion}

37
Latex/assets/tikzit/CausalGraph_joint.tikz
Unescape Escape View File

@ -0,0 +1,37 @@
\begin{tikzpicture}
\begin{pgfonlayer}{nodelayer}
\node [style=Red Box] (0) at (4, -1.5) {Will Terminate?};
\node [style=Red Box] (1) at (-4.25, -1.5) {Market Measures};
\node [style=emptyBox] (4) at (-6, -7.5) {Unobserved};
\node [style=Red Box] (5) at (2.5, 2) {Enrollment Status - Elapsed Duration};
\node [style=Red Box] (8) at (-5.75, -6) {Relationships of interest};
\node [style=emptyBox] (9) at (12.25, 6.25) {Fundamental Efficacy/Safety};
\node [style=emptyBox] (10) at (-0.25, 8) {Previously Observed Efficacy/Safety};
\node [style=emptyBox] (12) at (14, 0.25) {Currently Observed Efficacy/Safety};
\node [style=Gray Box] (13) at (7, -6) {Observed};
\node [style=Gray Box] (18) at (-6.5, 1.5) {Population};
\node [style=Gray Box] (20) at (-0.25, 5) {Decision to procced with Phase III};
\node [style=Gray Box] (21) at (0, -3.5) {Condition};
\end{pgfonlayer}
\begin{pgfonlayer}{edgelayer}
\draw [style=RightArrow] (18) to (5);
\draw [style=RightArrow] (18) to (1);
\draw [style=RightArrow] (9) to (10);
\draw [style=RightArrow] (10) to (20);
\draw [style=RightArrow] (9) to (12);
\draw [style=RightArrow] (12) to (0);
\draw [style=Light Arrow] (21) to (1);
\draw [style=Light Arrow] (21) to (0);
\draw [style=Light Arrow, in=-120, out=165, looseness=2.00] (21) to (18);
\draw [style=Light Arrow, in=180, out=180, looseness=2.75] (21) to (20);
\draw [style=Light Arrow, bend right=60, looseness=1.75] (21) to (9);
\draw [style=Light Arrow] (21) to (5);
\draw [style=RightArrow, in=-180, out=165, looseness=2.00] (1) to (20);
\draw [style=RightArrow] (20) to (5);
\draw [style=RightArrow, bend left] (20) to (0);
\draw [style=PurpleArrow] (1) to (0);
\draw [style=PurpleArrow] (1) to (5);
\draw [style=PurpleArrow] (5) to (0);
\draw [style=PurpleArrow, in=135, out=45, loop] (8) to ();
\end{pgfonlayer}
\end{tikzpicture}
Write Preview
Loading…
Cancel
Save