-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathseparability.tex
192 lines (149 loc) · 8.75 KB
/
separability.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
\documentclass{bioinfo}
\usepackage{latexsym, mathrsfs}
\usepackage{slantsc}
\usepackage{xspace}
\usepackage[titletoc,title]{appendix}
\usepackage{amsthm,amssymb}
\usepackage{subfigure} %{subfig} {subcaption}
\usepackage{multirow}
\usepackage[normalem]{ulem}
\usepackage{url}
\usepackage{multibib}
\newcites{app}{Supplementary References}
%\usepackage[usenames,dvipsnames]{color}
%\usepackage[linesnumbered,ruled,procnumbered]{algorithm2e}
%\usepackage{paralist}
%\usepackage{url}
%\usepackage{graphicx}
%\usepackage{floatflt}
%\usepackage{times}
%\usepackage{balance}
%\usepackage{spverbatim}
%\usepackage{hyperref}
%%%%%%% adjust spacing
\makeatletter
\renewcommand\section{\@startsection{section}{1}{\z@}%
{-8\p@ \@plus -4\p@ \@minus -4\p@}%
{6\p@ \@plus 4\p@ \@minus 4\p@}%
{\normalfont\large\bfseries\boldmath
\rightskip=\z@ \@plus 8em\pretolerance=10000 }}
\renewcommand\subsection{\@startsection{subsection}{2}{\z@}%
{-8\p@ \@plus -4\p@ \@minus -4\p@}%
{6\p@ \@plus 4\p@ \@minus 4\p@}%
{\normalfont\normalsize\bfseries\boldmath
\rightskip=\z@ \@plus 8em\pretolerance=10000 }}
\renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}%
{-4\p@ \@plus -4\p@ \@minus -4\p@}%
{-1.5em \@plus -0.22em \@minus -0.1em}%
{\normalfont\normalsize\bfseries\boldmath}}
\makeatother
\newcommand{\squishlist}{\begin{list}{$\bullet$}{\leftmargin=1.2em \topsep 1pt \itemsep -1pt \parsep 1pt}}
\newcommand{\squishend}{\end{list}}
%%%%%%%%%%% special formatting
\newcommand{\sampar}[1]{\vspace{3pt}\noindent{\bf #1}}
\newcommand{\topic}[1]{\vspace{1pt} \noindent \underline{\bf #1}}
\newcommand{\stitle}[1]{\vspace{0.5em}\noindent\textbf{#1}}
%%%%%%%%%%% Theorem templates
\newtheorem{theorem}{Theorem}
\newtheorem{lemma}{Lemma}
\newtheorem{conjecture}{Conjecture}
\newtheorem{example}{Example}
\newtheorem{formulation}{Problem}
\newtheorem{definition}{Definition}
\newtheorem{observation}[theorem]{\textbf{Observation}}
\newtheorem{claim}[lemma]{\textbf{Claim}}
\newtheorem{problem}{\textbf{Problem}}
%\theoremstyle{slplain}
%%%%%%%%%%%%% special math notation
\newcommand{\mm}{{\mathcal{M}}\xspace}
\newcommand{\ff}{{\mathcal{F}}\xspace}
\newcommand{\oo}{{\mathcal{O}}\xspace}
\newcommand{\sss}{{\mathcal{S}}\xspace}
\newcommand{\cc}{{\mathcal{C}}\xspace}
\newcommand{\lhat}{\mathit{L}\xspace}
%%%%%%%%%%%%%%% special words
\newcommand{\topk}{top-k\xspace}
\newcommand{\topthree}{top-3 }
\newcommand{\toptwenty}{top-20 }
\newcommand{\tophundred}{top-100 }
\newcommand{\topthousand}{top-1000 }
%\newcommand{\topthree}{{\sc T{\small OP-3}}\xspace}
\newcommand{\trans}{{\sc T{\small RANS\-FORMATION}}\xspace}
\newcommand{\genviz}{{\sc G{\small ENVISAGE}}\xspace}
\newcommand{\earlyT}{{\sc E{\small ARLY}S{\small TOP}}\xspace}
\newcommand{\sampling}{{\sc S{\small AMPLING}}\xspace} %\hspace{.2mm}
\newcommand{\traversal}{{\sc T{\small RAVERSAL}}\xspace}
\newcommand{\baseline}{{\sc Baseline}\xspace}
\newcommand{\early}{{\sc Early}\xspace}
\newcommand{\earlyOrder}{{\sc EarlyOrdering}\xspace}
\newcommand{\samp}{{\sc SampOnly}\xspace}
\newcommand{\sampOpt}{{\sc SampOpt}\xspace}
\newcommand{\horiz}{{\sc HorizSampOpt}\xspace}
\newcommand{\vertic}{{\sc VertSampOpt}\xspace}
\newcommand{\msig}{{\sc MSigDB}\xspace}
\newcommand{\lincs}{{\sc LINCS}\xspace}
%%%%%%%%%%%%% editing tools
\newcommand{\sinha}[1]{\noindent{\textcolor{blue}{[Saurabh: #1]}}}
\newcommand{\agp}[1]{\noindent{\textcolor{red}{[Aditya: #1]}}}
\newcommand{\agpins}[1]{\noindent{\textcolor{red}{#1}}}
\newcommand{\cb}[1]{\noindent{\textcolor{cyan}{[Charles: #1]}}}
\newcommand{\silu}[1]{\noindent{\textcolor{magenta}{[Silu: #1]}}}
\newcommand{\xsinha}[1]{\noindent{\textcolor{green}{[X SS: #1]}}}
\newcommand{\xagp}[1]{\noindent{\textcolor{green}{[X AP: #1]}}}
\newcommand{\xcb}[1]{\noindent{\textcolor{green}{[X CB: #1]}}}
\newcommand{\xsilu}[1]{\noindent{\textcolor{green}{[X SH: #1]}}}
\newcommand{\todo}[1]{\noindent{\textcolor{orange}{\textbf{[TODO]: #1}}}}
\newcommand{\needcite}[1]{\noindent{\textcolor{blue}{\textbf{[CITE] #1}}}}
\newcommand{\expn}[1]{\noindent{\textcolor{blue}{[Expand?]: #1}}}
\newcommand{\trim}[1]{\noindent{\textcolor{magenta}{[Reduce?]: #1}}}
\newcommand{\papertext}[1]{\noindent{\textcolor{blue}{PAPER: #1}}}
\newcommand{\cut}[1]{}
\newcommand{\eat}[1]{}
\newcommand{\tr}[1]{}
\newcommand{\tleftout}[1]{}
\newcommand{\new}[1]{\noindent{\textcolor{red}{#1}}}
%\newcommand{\tleftout}[1]{\noindent{\textcolor{blue}{#1}}}
%%%%%%%%% old supplementary references
%\newcommand{\timetbl}{Supplementary Table 1}
%\newcommand{\fpstbl}{Supplementary Table 2}
%\newcommand{\transfig}{Supplementary Figure 1}
%\newcommand{\bruterocchioratio}{Supplementary Figure 2(a)}
%\newcommand{\bruterocchioscore}{Supplementary Figure 2(b)}
%\newcommand{\histogramdiff}{Supplementary Figure 3}
%\newcommand{\travnote}{Supplementary Note 1}
\copyrightyear{2018} \pubyear{2018}
\access{} %Advance Access Publication Date: Day Month Year
\appnotes{Original Paper} %Manuscript Category
\begin{document}
\firstpage{1}
\subtitle{} %Subject Section
\title[Genvisage]{G{\large ENVISAGE}: Rapid Identification of Discriminative Feature Pairs for Genomic Analysis}
\author[Huang \textit{et~al}.]{Silu Huang\,$^{\text{\sfb 1}\dagger}$, Charles Blatti\,$^{\text{\sfb 2}\dagger}$, Saurabh Sinha\,$^{\text{\sfb 1,2}\ast}$, and Aditya Parameswaran\,$^{\text{\sfb 1,2}\ast}$ }
\address{$^{\text{\sf 1}}$Department of Computer Science, University of Illinois Urbana-Champaign, Urbana, IL, 61801, USA and \\
$^{\text{\sf 2}}$Institute of Genomic Biology, University of Illinois Urbana-Champaign, Urbana, IL, 61801, USA.}
\corresp{
$^\ast$To whom correspondence should be addressed. \\
$^\dagger$These authors contributed equally to this work.
}
\history{} %Received on XXXXX; revised on XXXXX; accepted on XXXXX
\editor{} %Associate Editor: XXXXXXX
%\abstract{
%A common but critical task in biological data analysis is the {\em separability} task: finding features that explain the difference between two different classes of objects with high dimensional feature representations, such as genes described by their functional annotations, or gene signatures described by their transcriptomic profiles. We develop an interactive data exploration tool called \genviz for this task that rapidly identifies discriminative feature pairs and outputs the corresponding visualizations. Since quickly finding top-k feature pairs is computationally challenging, especially when the numbers of objects and features are large, we propose a suite of optimizations to make \genviz more responsive, and demonstrate that our optimizations lead to a 400X speedup over competitive baselines for multiple biological data sets. With this speedup, \genviz enables the exploration of more datasets and alternative hypotheses in an interactive fashion. Finally, we apply \genviz to uncover pairs of genes whose transcriptomic responses more significantly discriminate treatments of a given drug.}
\abstract{\textbf{Motivation:} A common but critical task in biological data analysis is the {\em separability} task: finding features that explain the difference between two different classes of objects with high dimensional feature representations, such as genes described by their functional annotations or gene signatures described by their transcriptomic profiles.\\
\textbf{Results:} We develop an interactive data exploration tool called \genviz for this task that rapidly identifies discriminative feature pairs and outputs the corresponding visualizations. Since quickly finding top-k feature pairs is computationally challenging, especially when the numbers of objects and features are large, we propose a suite of optimizations to make \genviz more responsive, and demonstrate that our optimizations lead to a 400X speedup over competitive baselines for multiple biological data sets. With this speedup, \genviz enables the exploration of more datasets and alternative hypotheses in an interactive fashion. Finally, we apply \genviz to uncover pairs of genes whose transcriptomic responses more significantly discriminate treatments of different chemotherapy drugs.\\
\textbf{Availability:} Free webserver at \url{http://genvisage.knoweng.org:443/} with source code at \url{https://github.com/KnowEnG/Genvisage} \\
%\textbf{Contact:} \href{mailto:[email protected]}{[email protected]}\\
\textbf{Supplementary information:} Supplementary data are available at \textit{Bioinformatics} online.}
\maketitle
\input{body_short.tex}
{\scriptsize
\bibliographystyle{unsrt}
\bibliography{ref}
}
\clearpage
\begin{appendices}
\input{appendix.tex}
\bibliographystyleapp{unsrt}
\bibliographyapp{ref}
\end{appendices}
\end{document}