Merge https://git.overleaf.com/5d5e7765e8f5a437657dc755

Author: AldisiRana

content/biological.tex

@@ -25,7 +25,7 @@ \subsection{Targets}
 \subsection{Side Effects}
 
 Even though drugs are taken for their therapeutic effects, they also have the potential risk of being harmful.
-Since drugs’ effect inside the body are not only limited to their intended targets, sometimes, they can cause unintended medical reactions in the body.
+Since drugs’ effect inside the body is not only limited to their intended targets, sometimes, they can cause unintended medical reactions in the body.
 These are known as a side effects, or adverse drug events~\cite{pourpak_understanding_2008}.
 While their causes generally lack a mechanistic understanding, some intrinsic risk factors have been suggested for their developments such as age, gender, weight, genetics, and state of health.
 They could also be affected by extrinsic factors like the dosage of the drug, the route of administration, or taking multiple drugs at the same time~\cite{pourpak_understanding_2008}.

content/methods.tex

@@ -53,7 +53,7 @@ \section{Experimental Runs}
 These models were used based on their application and performance in~\cite{yue_graph_2019}.
 The best model was selected based on the collective evaluation results between all optimized models.
 
-\subsection{\ac{NRL} Models}
+\subsection{NRL Models}
 
 To choose the best prediction model for the network, six different models from three categories were selected and tested.
 \ac{HOPE} and \ac{GraRep} were chosen from the matrix factorization approaches, \ac{LINE} and \ac{SDNE} were selected from the deep learning methods, and from random walk approaches, DeepWalk and node2vec were chosen.

content/results.tex

@@ -201,7 +201,7 @@ \subsection{Predicting the Phenotypes for a Drug}
 \subsection{Predicting the Drugs for a Phenotype}
 
 The model can also be used to predict chemicals that can be associated with phenotypes.
-Here, it used to predict chemicals that might best affect \ac{PD}, a progressive neurodegenerative disorder that affects motor and non-motor functions in variable degrees~\cite{jankovic_parkinsons_2008}.
+Here, it is used to predict chemicals that might best affect \ac{PD}, a progressive neurodegenerative disorder that affects motor and non-motor functions in variable degrees~\cite{jankovic_parkinsons_2008}.
 Common motor features of \ac{PD} are tremors, rigidity, slowness (bradykinesia), and impaired balance.
 Other \ac{PD} symptoms include cognitive impairment and abnormal neurological behaviors~\cite{jankovic_parkinsons_2008}.
 The top ten predictions of chemicals associated with \ac{PD} are shown in Table~\ref{tab:phenotype_drug}.
@@ -268,7 +268,7 @@ \subsection{Predicting the Drugs for a Phenotype}
     \caption[Quetiapine's path subgraph in Parkinson's disease]{\label{fig:parkinson_quetiapine} A subgraph showing the shortest relation paths between quetiapine and \ac{PD}.}
 \end{figure}
 
-Two high-scoring chemicals chemicals, pregabalin and ziprasidone, have been shown to cause or worsen symptoms of \ac{PD}~\cite{perez_lloret_pregabalin-induced_2009, younce_systematic_2019}, while the rest of the predicted chemicals do not have any studies that prove their association with \ac{PD}.
+Two high-scoring chemicals, pregabalin and ziprasidone, have been shown to cause or worsen symptoms of \ac{PD}~\cite{perez_lloret_pregabalin-induced_2009, younce_systematic_2019}, while the rest of the predicted chemicals do not have any studies that prove their association with \ac{PD}.
 The positive controls for this association are shown in Table~\ref{tab:ps_PD}.
 These four drugs (i.e., selegiline, aripiprazole, ropinirole, and clomipramine) are already used for treating \ac{PD} and are indicated as such in DrugBank, which is why their association with the disease exists in the network and they are expected to be predicted by the model.