Skip to content
Snippets Groups Projects
Commit 85b1a786 authored by Vermaat's avatar Vermaat
Browse files

Merge 235 from trunk. 

git-svn-id: https://humgenprojects.lumc.nl/svn/mutalyzer/branches/refactor-mutalyzer-branch@241 eb6bd6ab-9ccd-42b9-aceb-e2899b4a52f1
parents 2595121b 30b16f81
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
doc/Presentation_24-02-11_HumGen_Mutalyzer2/leftover.txt deleted 100644 → 0
+ 0
− 271
View file @ 2595121b
% - Mutalyzer 1.0.4
\begin{slide}
\slideheading{Mutalyzer 1.0.4}
\begin{itemize}
\item Developed in over four years by multiple people.
\item Originally a command line program.
\item Web interface added later.
\end{itemize}
\vfill
\end{slide}
% - Design flaws:
% - Nomenclature rules interwoven with the code.
% - No modularity (reuse of code is very hard).
% - Reference sequence parsing not abstracted.
% - HTML output interwoven with the code.
\begin{slide}
\slideheading{Mutalyzer 1.0.4}
Design flaws:
\begin{itemize}
\item Nomenclature rules interwoven with the code.
\item HTML output interwoven with the code.
\item No modularity (reuse of code is very hard).
\item Reference sequence parsing not abstracted.
\end{itemize}
\vfill
\end{slide}
% - Implementation flaws:
% - Inheritance of types (del on DNA -> del on PROT).
% - Disambiguation not general.
% - Support up/downstream exons.
% - Speed
\begin{slide}
\slideheading{Mutalyzer 1.0.4}
Implementation flaws:
\begin{itemize}
\item Inheritance of types (del on DNA -> del on PROT).
\item Disambiguation not general.
\item Support up/downstream exons.
\item Nothing was ever redesigned, only wrapped in loops.
\begin{itemize}
\item Debugging, altering code made impossible.
\item Speed drastically deminished.
\end{itemize}
\end{itemize}
\vfill
\end{slide}
% - Programming flaws:
% - Excessive usage of exceptions.
% - Incomprehensible error messages.
% - Poor documentation.
\begin{slide}
\slideheading{Mutalyzer 1.0.4}
Programming flaws:
\begin{itemize}
\item Excessive usage of exceptions.
\item Incomprehensible error messages.
\item Poor documentation.
\end{itemize}
\vfill
\end{slide}
% - Feature requests:
% - Extension of HGVS nomenclature rules.
% - Support for other reference files (LRG)
% - Programmatic access to internal functions.
% - Solving all problems mentioned above.
% - Since the nomenclature has changed, a rewrite was in order.
%
\begin{slide}
\slideheading{Mutalyzer 1.0.4}
Feature requests:
\begin{itemize}
\item Solving all problems mentioned above.
\item Support for other reference files (LRG).
\item Programmatic access to internal functions.
\end{itemize}
Since the HGVS nomenclature rules were changed in the mean time, and the
language was no longer regular (but context free), the only possible couse of
action was a complete redesign.
\vfill
\end{slide}
% - Preparations for version 2.0
% - Gathering and archiving all old versions (for comparison).
% - Setting up a version control repository.
% - Talking for months.
% - Figuring out what the HGVS language is.
% - Formalising that language (BNF).
% - Semantic rules.
% - Chopping everything up in functional modules.
% - Designing interfaces (web, webservice, command line, etc.)
\begin{slide}
\slideheading{Preparing for a new version}
\begin{itemize}
\item Setting up a version control repository.
\item Gathering all old versions and putting then under version control.
\begin{itemize}
\item Critical bugfixes until there is a new version.
\item Easy to search and track changes.
\item Point of reference for the new version.
\end{itemize}
\item Talking for months.
\begin{itemize}
\item Figuring out what the HGVS language is.
\item Formalising that language (BNF).
\item Semantic rules.
\end{itemize}
\item Chopping everything up in functional modules.
\item Designing interfaces (web, webservice, command line, etc.).
\end{itemize}
\vfill
\end{slide}
%
% - Then finally, after months of talking and drawing with pencil and paper..
% - Implementing the modules.
% - Implementing the interfaces.
\begin{slide}
\slideheading{Mutalyzer 2.0}
Then finally, after months of talking and drawing with pencil and paper..
\begin{itemize}
\item Implementing the modules.
\item Implementing the interfaces.
\end{itemize}
\vfill
\end{slide}
%
% - Mutalyzer 2.0
% - Core functionalities.
% - Webservices.
% - ...
\begin{slide}
\slideheading{TAL}
\begin{lstlisting}[language = HTML, caption = {TAL example}]
<table class = "raTable">
<tr>
<td>Number</td>
<td>Start (g.)</td>
<td>Stop (g.)</td>
<td>Start (c.)</td>
<td>Stop (c.)</td>
</tr>
<tr tal:repeat = "i exonInfo">
<td tal:content = "repeat/i/number"></td>
<td tal:repeat = "j i" tal:content = "j"></td>
</tr>
</table>
\end{lstlisting}
When we give a list of exon coordinates, a table is generated.
\vfill
\end{slide}
\begin{slide}
\slideheading{BNF}
\begin{lstlisting}[language = BNF, caption = {Abstract HGVS nomenclature}]
TransVar -> `_v' Number
ProtIso -> `_i' Number
GeneSymbol -> `(' Name (TransVar | ProtIso)? `)'
\end{lstlisting}
\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}
\bt{(CDKN2A\_v001)}
\begin{lstlisting}[caption = {Python object}]
Gene.GeneSymbol = CDKN2A
Gene.TransVar = 001
\end{lstlisting}
\bt{(CDKN2A\_i002)}
\begin{lstlisting}[caption = {Python object}]
Gene.GeneSymbol = CDKN2A
Gene.ProtIso = 002
\end{lstlisting}
\vfill
\end{slide}
\begin{slide}
\slideheading{Comparison to the old version (1.0.4)}
\renewcommand{\arraystretch}{0.99}
\begin{tabular}{l|c|c}
& Mutalyzer 1.0.4 & Mutalyzer 2.0\\
\hline
Disambiguation & $\pm$ & $++$\\
Complex variants & $--$ & $++$\\
Protein description & $\pm$ & $+$\\
Up / downstream descriptions & $--$ & $++$\\
Comprehensible error messages & $-$ & $++$\\
Using a protein reference & $\pm$ & $--$\\
Batch checkers & $\pm$ & $++$\\
GenBank uploader & $+$ & $++$\\
Position conversion & $--$ & $++$\\
Programmatic access & $--$ & $++$\\
Other organisms / organelles & $\pm$ & $++$\\
\end{tabular}
\vfill
\end{slide}
\begin{slide}
\slideheading{Comparison to the old version (1.0.4): runtime}
\begin{center}
\colorbox{white} {
\includegraphics[scale = 0.65]{genes}
}
\end{center}
A $229\times$ speedup was measured (from almost $12min$ to about $3s$).
\vfill
\end{slide}
\begin{slide}
\slideheading{Comparison to the old version (1.0.4): code}
\begin{tabular}{l|r|r}
& Mutalyzer 1.0.4 & Mutalyzer 2.0\\
\hline
Total (lines) & $7,\!752$ & $11,\!396$\\
Total (bytes) & $365,\!736$ & $390,\!316$\\
Minimised (lines) & $5,\!102$ & $4,\!320$\\
Minimised (bytes) & $232,\!611$ & $156,\!803$\\
Percentage of code (lines) & $66\%$ & $38\%$\\
Percentage of code (bytes) & $64\%$ & $42\%$
\end{tabular}
\bigskip
\bigskip
The total amount of \emph{source code} in Mutalyzer~2.0 is $107\%$ of that in
Mutalyzer~1.0.4, but the amount of \emph{program code} is only $67\%$.
\vfill
\end{slide}
\begin{slide}
\slideheading{Scalability: runtime with increasing complexity}
\begin{center}
\colorbox{white} {
\includegraphics[scale = 0.65]{allele}
}
\end{center}
The overhead ($\pm 2.5s$) is due to loading the reference sequence.
\vfill
\end{slide}
mutalyzer/variantchecker.py
+ 5
− 0
View file @ 85b1a786
......@@ -1055,6 +1055,11 @@ def _add_transcript_info(mutator, transcript, output):
cds_original = Bio.Seq.reverse_complement(cds_original)
cds_variant = Bio.Seq.reverse_complement(cds_variant)
if not util.is_dna(cds_original):
output.addMessage(__file__, 4, 'ENODNA',
'Invalid letters in reference sequence.')
return
if '*' in cds_original.translate(table=transcript.txTable)[:-1]:
output.addMessage(__file__, 3, 'ESTOP',
'In frame stop codon found.')
......
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