Skip to content
Snippets Groups Projects
Commit 77325486 authored by Laros's avatar Laros
Browse files

Added a presentation that covers the description extractor. 


git-svn-id: https://humgenprojects.lumc.nl/svn/mutalyzer/trunk@608 eb6bd6ab-9ccd-42b9-aceb-e2899b4a52f1
parent 3185a420
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
Showing
with 687 additions and 0 deletions
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/Makefile 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/Makefile
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/beamerthemelumc.sty 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/beamerthemelumc.sty
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/gen2phen_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/gen2phen_logo.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/lgtc_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/lgtc_logo.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/lstBNF.tex 0 → 120000
+ 1
− 0
View file @ 77325486
../Presentation_24-02-11_HumGen_Mutalyzer2/lstBNF.tex
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/lumc_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/lumc_logo.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/lumc_logo_small.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/lumc_logo_small.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/nbic_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/nbic_logo.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/ngi_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/ngi_logo.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/nwo_logo_en.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/nwo_logo_en.eps
\ No newline at end of file
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/presentation.tex 0 → 100644
+ 676
− 0
View file @ 77325486
\documentclass[slidestop]{beamer}
\title{Extracting HGVS descriptions}
\providecommand{\myConference}{Work discussion}
\providecommand{\myDate}{Thursday, 24 February 2011}
\author{Jeroen F. J. Laros}
\providecommand{\myGroup}{Leiden Genome Technology Center}
\providecommand{\myDepartment}{Department of Human Genetics}
\providecommand{\myCenter}{Center for Human and Clinical Genetics}
\providecommand{\lastCenterLogo}{
\raisebox{-0.1cm}{
\includegraphics[height = 1cm]{lgtc_logo}
%\includegraphics[height = 0.7cm]{ngi_logo}
}
}
\providecommand{\lastRightLogo}{
%\includegraphics[height = 0.7cm]{nbic_logo}
%\includegraphics[height = 0.8cm]{nwo_logo_en}
\hspace{1.5cm}\includegraphics[height = 0.7cm]{gen2phen_logo}
}
\usetheme{lumc}
\usepackage{ifthen}
\input{lstBNF}
\begin{document}
\newcommand{\algorithmexample}[1]{
\begin{figure}[]
\begin{center}
\fbox{
\setlength{\unitlength}{1pt}
\linethickness{3pt}
\begin{picture}(300, 60)(0, 0)
\put(0, 10){\line(1, 0){30}} % Observed sequence.
\put(30, 10){\color{red}\line(1, 0){240}\color{white}} % Change.
\put(270, 10){\line(1, 0){30}}
\put(0, 14){{\scriptsize observed}}
\put(0, 40){\line(1, 0){30}} % Reference sequence.
\put(30, 40){\color{green}\line(1, 0){240}\color{white}} % Change.
\put(270, 40){\line(1, 0){30}}
\put(0, 46){{\scriptsize reference}}
\put(30, 30){{\scriptsize $8$}}
\put(270, 30){{\scriptsize $98$}}
\ifthenelse{\equal{#1}{1}}{
\drawcurve(50, 40)(55, 35)(155, 25)(255, 15)(260, 10)
\drawcurve(260, 40)(255, 35)(155, 25)(55, 15)(50, 10)
}{}
\ifthenelse{#1>1}{
\put(50, 10){\line(1, 0){210}} % Inv.
\put(50, 40){\line(1, 0){210}} % Inv.
}{}
\ifthenelse{#1>2}{
\put(35, 10){\line(1, 0){10}}
\put(35, 40){\line(1, 0){10}}
}{}
\end{picture}
}
\end{center}
\caption{How would a human do it?}
\end{figure}
}
% This disables the \pause command, handy in the editing phase.
%\renewcommand{\pause}{}
% Make the title page.
\bodytemplate
% First page of the presentation.
\section{Introduction}
\begin{frame}
\frametitle{Mutalyzer}
A curational tool for \emph{Locus Specific Mutation Databases} (LSDBs).
\pause
\bigskip
Variant nomenclature checker applying \emph{Human Genome Variation Society}
(HGVS) guidelines.
\begin{itemize}
\item Is the syntax of the variant description valid?
\item Does the reference sequence exist?
\item Is the variant possible on this reference sequence?
\item Is this variant description the recommended one?
\end{itemize}
\bigskip
\pause
Basic effect prediction.
\begin{itemize}
\item Is the description of the transcript product as expected?
\item Is the predicted protein as expected?
\end{itemize}
\end{frame}
\begin{frame}
\frametitle{Mutalyzer}
Nowadays Mutalyzer is a vital part of LOVD version 3.
\bigskip
\pause
Make a reference sequence (configure new gene):
\begin{itemize}
\item Given a gene symbol, make a slice of a chromosome.
\item Receive information on the transcripts and genes in a genomic
reference sequence.
\end{itemize}
\medskip
\pause
Mapping variants:
\begin{itemize}
\item Find which transcript is affected.
\item Map variants to the genome and vice versa.
\item Lift a description over to an other transcript.
\end{itemize}
\medskip
\pause
Curating submissions:
\begin{itemize}
\item Checking the syntax.
\item Checking the variant description.
\end{itemize}
\end{frame}
\section{HGVS nomenclature}
\begin{frame}
\frametitle{HGVS descriptions}
A simple variant:
\bt{NM\_002001.2:c.25A>T}
\bigskip
\pause
\begin{table}[]
\begin{center}
\begin{tabular}{c|l}
Token & meaning \\
\hline
\bt{NM\_002001.2} & Reference sequence and version. \\
\bt{c.} & Coordinate system. \\
\bt{25} & Position within a coordinate system. \\
\bt{A>T} & Variant (substitution). \\
\end{tabular}
\end{center}
\caption{A simple variant description.}
\end{table}
\bigskip
\pause
Combine simple variants to complex ones:
\bt{NM\_002001.2:c.[25A>T;100del]}
\end{frame}
\begin{frame}[fragile]
\frametitle{HGVS syntax}
\pause
Definition of a gene symbol.
\begin{lstlisting}[language = BNF, caption = {Abstract HGVS nomenclature}]
TransVar -> `_v' Number
ProtIso -> `_i' Number
GeneSymbol -> `(' Name (TransVar | ProtIso)? `)'
\end{lstlisting}
\bigskip
\pause
Gene name and optionally a transcript or isoform number.
\begin{lstlisting}[caption = {HGVS nomenclature in Python}]
TransVar = Suppress("_v") + Number("TransVar")
ProtIso = Suppress("_i") + Number("ProtIso")
GeneSymbol = Suppress('(') + \
Group(Name("GeneSymbol") + \
Optional(TransVar ^ ProtIso))("Gene") + \
Suppress(')')
\end{lstlisting}
\end{frame}
\begin{frame}
\frametitle{HGVS semantics}
There are a few guidelines for describing variants:
\begin{itemize}
\item Always use the most 5' variant description.
\item Use the shortest description.
\end{itemize}
\bigskip
\pause
There are no guidelines on \emph{how} to do this.
\bigskip
Example: we observe a change from \bt{CCCCCCC} to \bt{CACACAC}.
\begin{itemize}
\item \bt{2\_6\color{yellow}delins\color{white}ACACA}
\item \bt{[2C\color{yellow}>\color{white}A;4C\color{yellow}>\color{white}A;6C\color{yellow}>\color{white}A]}
\item \bt{[1\_2\color{yellow}ins\color{white}A;3\_6\color{yellow}delins\color{white}ACA]}
\item \ldots
\end{itemize}
\end{frame}
\begin{frame}
\frametitle{Variants are not ``inherited''}
Silent mutations for example.
\bigskip
\pause
A double frameshift:
\bt{NM\_002001.2:c.[10del;22\_23del]}
\bt{NP\_001992.1:p.Ala4\_Pro7delinsProTrpAsn}
\bigskip
\pause
A complex variant that leads to a simple protein change:
\bt{NM\_002001.2:[c.10\_12delinsAAA;102G>A]}
\bt{NP\_001992.1:p.Ala4Lys}
\bigskip
\pause
An insertion that affects two codons:
\bt{NM\_002001.2:c.10\_11insTTT}
\bt{NP\_001992.1:p.Ala4delinsValSer}
\end{frame}
\begin{frame}
\frametitle{Problem description}
Verifying the validity of a variant description is not enough:
\begin{itemize}
\item Both \bt{5\_7delinsATA} and \bt{[6G>A;7C>A]} are valid.
\item We want one representation.
\end{itemize}
\bigskip
\pause
We need something that:
\begin{itemize}
\item Accepts any description to modify a reference sequence.
\item Compares the reference and the modified sequence to make a
description.
\end{itemize}
\bigskip
\pause
A description extractor.
\end{frame}
\section{Extracting descriptions}
\begin{frame}
\frametitle{A ``human'' way of finding a description}
Observation:
\begin{itemize}
\item There is always a default way of describing a variant (\bt{delins}).
\item A \bt{delins} may be split in smaller parts.
\end{itemize}
\bigskip
\pause
Outline:
\begin{itemize}
\item Find the \emph{area of change}.
\item Describe this as a \bt{delins}.
\item Find the largest overlap in this area of change, splitting the area
in two.
\item Describe the two sub areas, and see whether this description is
smaller than the one we have.
\end{itemize}
\end{frame}
\begin{fframe}
\frametitle{Outline of the algorithm}
\only<1>{\algorithmexample{0}}
\only<2>{\algorithmexample{1}}
\only<3>{\algorithmexample{2}}
\only<4>{\algorithmexample{3}}
\bt{8\_98\color{yellow}delins\color{white}AGATGCGATAGATTAGCTATATAGGATCG\ldots}
\onslide<3->{\bt{[8\_12\color{yellow}delins\color{white}AGATG;13\_96\color{yellow}inv\color{white};97\_98\color{yellow}delins\color{white}TG]}}
\onslide<4->{\bt{[8G\color{yellow}>\color{white}A;12C\color{yellow}>\color{white}G;13\_96\color{yellow}inv\color{white};97\_98\color{yellow}delins\color{white}TG]}}
\vfill
\end{fframe}
\begin{fframe}
\frametitle{Finding common sub strings}
How would a computer do it?
\begin{table}[]
\begin{center}
\begin{tabular}{l|lllllll}
& \bt{A} & \bt{T} & \bt{G} & \bt{A} & \bt{G} & \bt{C} & \bt{G} \\
\hline
\bt{A} & \onslide<2>{\color{red}}1 & 0 & 0 &
\onslide<3>{\color{gray}}1 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 \\
\bt{T} & 0 & \onslide<2>{\color{red}}2 & 0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 \\
\bt{C} & 0 & 0 & 0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}1 &
\onslide<3>{\color{gray}}0 \\
\bt{A} & \onslide<3>{\color{gray}}1 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}1 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 \\
\bt{G} & \onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}1 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}2 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}1 \\
\bt{C} & \onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}3 &
\onslide<3>{\color{gray}}0 \\
\bt{A} & \onslide<3>{\color{gray}}1 & \onslide<3>{\color{gray}}0 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}1 &
\onslide<3>{\color{gray}}0 & \onslide<3>{\color{gray}}0 & 0 \\
\end{tabular}
\end{center}
\caption{LCS dynamic programming.}
\end{table}
\only<2>{Reusing partial solutions.}
\only<3>{Reusing parts of the matrix.}
\vfill
\end{fframe}
\section{Results}
\begin{frame}
\frametitle{Protein descriptions}
Input:
\bt{NM\_002001.2:n.[109G>T;139G>T;159del]}
\bigskip
\pause
Old:
\bt{NM\_002001.2:n.[109G>T;139G>T;159del]}
\bt{NM\_002001.2:p.?}
\bigskip
\pause
New:
\bt{NM\_002001.2:n.[109G>T;139G>T;159del]}
\bt{NM\_002001.2:p.[Ala4Ser;Ala14Ser;Asp21Metfs*4]}
\end{frame}
\begin{frame}
\frametitle{Protein descriptions (2)}
Input:
\bt{NM\_002001.2:n.[159del;162\_163del]}
\bigskip
\pause
Old:
\bt{NM\_002001.2:n.[159del;162\_163del]}
\bt{NM\_002001.2:p.?}
\bigskip
\pause
New:
\bt{NM\_002001.2:n.[159del;162\_163del]}
\bt{NM\_002001.2:p.Asp21\_Val22delinsSer}
\end{frame}
\begin{frame}
\frametitle{Combining variants}
Input ($110$ and $111$ have the same nucleotide):
\bt{NM\_002001.2:n.[109del;111del]}
\bigskip
\pause
Old:
\bt{NM\_002001.2:n.[109del;111del]}
\bt{NM\_002001.2:p.?}
\bigskip
\pause
New:
\bt{NM\_002001.2:n.109\_110del}
\bt{NM\_002001.2:p.Ala4Hisfs*27}
\end{frame}
\begin{frame}
\frametitle{Splitting variants}
Input:
\bt{NM\_002001.2:c.40\_50delinsTCCTTACTGTG}
\bigskip
\pause
Old:
\bt{NM\_002001.2:n.139\_149delinsTCCTTACTGTG}
\bt{NM\_002001.2:p.Ala14\_Phe17delinsSerLeuLeuCys}
\bigskip
\pause
New:
\bt{NM\_002001.2:n.[139G>T;149T>G]}
\bt{NM\_002001.2:p.[Ala14Ser;Phe17Cys]}
\end{frame}
\begin{frame}
\frametitle{Comparing reference sequences}
DMD Dp71ab vs. DMD Dp71b:
\bigskip
Input:
\bt{NM\_004018.2} and \bt{NM\_004016.2}
\bigskip
\pause
Output:
\bt{1097\_1098insTCCCGTTACTCTGATCAACTTCTGGCCAGT\ldots}
\bigskip
Interpretation:
This is an exon not present in Dp71ab.
\end{frame}
\begin{frame}
\frametitle{Old vs. new transcripts}
DMD Dp71ab old vs. new:
\bigskip
Input: \bt{NM\_004018.2} and \bt{NM\_004018.1}
\bigskip
Output: \bt{[3308A>G;4288A>G]}
\bigskip
\bigskip
\bigskip
\pause
FCER1A old vs. new:
\bigskip
Input: \bt{NM\_002001.1} and \bt{NM\_002001.2}
\bigskip
Output: \bt{1\_7del}
\end{frame}
\begin{frame}
\frametitle{Old vs. new transcripts (2)}
FCER2 old vs. new:
\bigskip
Input:
\bt{NM\_002002.1} and \bt{NM\_002002.4}
\bigskip
\pause
Output:
\bt{[720C>T;903A>G;930T>C;1019C>A; \\
1401\_1402insACACCCCAACAGCACCCTCTCCAGATGAGAGT\ldots; \\
1478del;1529\_1530insTCCCACATTTGTCCCCTTCTTGGA\ldots]}
\smallskip
\pause
vice versa:
\bt{[720T>C;903G>A;930C>T;1019A>C;1402\_1464del; \\
1540dup;1592\_1620del]}
\end{frame}
\begin{frame}
\frametitle{Limitations}
mtDNA reference vs. isolate K422 mitochondrion
\medskip
Input: \bt{NC\_012920.1} and \bt{JX266268.1}
\medskip
\pause
Output:
\bt{
[73A>G;194C>T;249del;263A>G;310delinsCTC;489T>C; \\
750A>G;1438A>G;1715C>T;2231\_2232dup;2706A>G; \\
3107del;3552T>A;4715A>G;4769A>G;6026G>A;7028C>T; \\
7196C>A;7999T>C;8508A>G;8584G>A;8701A>G;8860A>G; \\
9540T>C;9545A>G;10398\_10400delinsGCT;10873T>C; \\
11719G>A;11914G>A;11969G>A;12672A>G;12705C>T; \\
13263A>G;14318T>C;14766C>T;14783T>C;15043G>A; \\
15204T>C;15301G>A;15326A>G;15487A>T;15968T>C; \\
16129G>A;16223C>T;16298T>C;16327C>T;16519T>C]
}
\medskip
\pause
Runtime: $\pm20$ minutes, Memory: $4$G.
\end{frame}
\section{Optimisation}
\begin{frame}
\frametitle{Accuracy vs. speed}
\begin{tabular}{l@{\ \ $\Rightarrow$\ \ }l}
\bt{AGAGGACG} & \bt{AG AG GA CG} \\
\bt{GAGGACA} & \bt{GA AG GG GA AC CA}
\end{tabular}
\pause
\begin{table}
\begin{center}
\begin{tabular}{l|llll}
& \bt{A} & \bt{A} & \bt{G} & \bt{C} \\
& \bt{G} & \bt{G} & \bt{A} & \bt{G} \\
\hline
\bt{GA} & 0 & 0 & 1 & 0 \\
\bt{AG} & 1 & \onslide<3>{\color{red}}1 & 0 & 0 \\
\bt{GG} & 0 & 0 & 0 & 0 \\
\bt{GA} & 0 & 0 & \onslide<3>{\color{red}}2 & 0 \\
\bt{AC} & 0 & 0 & 0 & 0 \\
\bt{CA} & 0 & 0 & 0 & 0 \\
\end{tabular}
\end{center}
\caption{Rough method to find large strings.}
\end{table}
\onslide<3>{We make a ``knight move''.}
\end{frame}
\begin{frame}
\frametitle{Accuracy vs. speed(2)}
\begin{minipage}[t]{0.45\textwidth}
\begin{table}[]
\begin{center}
\begin{tabular}{l|llll}
& \bt{A} & \bt{A} & \bt{G} & \bt{C} \\
& \bt{G} & \bt{G} & \bt{A} & \bt{G} \\
\hline
\bt{GA} & 0 & 0 & 1 & 0 \\
\bt{AG} & 1 & 1 & 0 & 0 \\
\bt{GG} & 0 & 0 & 0 & 0 \\
\bt{GA} & 0 & 0 & 2 & 0 \\
\bt{AC} & 0 & 0 & 0 & 0 \\
\bt{CA} & 0 & 0 & 0 & 0 \\
\end{tabular}
\end{center}
\caption{``Zoom out'' $k = 2$.}
\end{table}
\end{minipage}
\hfill
\begin{minipage}[t]{0.45\textwidth}
\begin{table}[]
\begin{center}
\begin{tabular}{l|ll}
& \bt{A} & \bt{G} \\
& \bt{G} & \bt{G} \\
& \bt{A} & \bt{A} \\
\hline
\bt{GAG} & 0 & 0 \\
\bt{AGG} & 0 & 0 \\
\bt{GGA} & 0 & 1 \\
\bt{GAC} & 0 & 0 \\
\bt{ACA} & 0 & 0 \\
\end{tabular}
\end{center}
\caption{``Zoom out'' $k = 3$.}
\end{table}
\end{minipage}
\pause
We find all common sub strings larger than $k$.
\pause
The length of these strings are at least $\ell k$ and at most
$\ell k + (k - 1)$ long.
\end{frame}
\section{Conclusions}
\begin{frame}
\frametitle{We are getting there}
Extracting descriptions is feasible.
\pause
\begin{itemize}
\item Guarantees the same description for the same variant, no matter how
it is described by the user.
\item Usable for comparing reference sequences.
\pause
\begin{itemize}
\item Real lift over.
\end{itemize}
\end{itemize}
\bigskip
\pause
Extracting descriptions is practical.
\begin{itemize}
\item By ``zooming out'', we can meet the memory requirements.
\begin{itemize}
\item $4$G to less than a megabyte.
\end{itemize}
\item By ``zooming out'', we can meet the processing requirements.
\begin{itemize}
\item mtDNA test: $20$ minutes to under one second.
\end{itemize}
\end{itemize}
\end{frame}
\section{Questions?}
\lastpagetemplate
\begin{frame}
\begin{center}
Acknowledgements:
\bigskip
\bigskip
Martijn Vermaat
Ivo Fokkema
Peter Taschner
Johan den Dunnen
\end{center}
\end{frame}
\end{document}
doc/Presentation_26-09-12_Lab-J_Descriprion_extract/ul_logo.eps 0 → 120000
+ 1
− 0
View file @ 77325486
/local/projects/presentation/trunk/ul_logo.eps
\ No newline at end of file
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment