Commit 39b0587d authored by Jens Behrmann's avatar Jens Behrmann

Update ReadMe.txt

parent 2225594d
##########################################################
ReadMe for the code of the article "Deep Learning for
Tumor Classification in Imaging Mass Spectrometry" by
Jens Behrmann, Christian Etmann, Tobias Boskamp,
Rita Casadonte, Jörg Kriegsmann, Peter Maass.
Should any questions arise, we would be happy to help you:
jensb@math.uni-bremen.de
cetmann@math.uni-bremen.de
##########################################################
The provided files are structured in the following way:
1) CNNs_Bioinformatics
This is where executable MATLAB scripts are
stored which reproduce the results found in the
article.
2) MSClassifyLib
This folder contains parts of a larger library
developed at the Center for Industrial Mathematics,
Bremen. It provides routines for the analysis of mass
spectrometry data and is required for the execution
of the scripts in the CNNs_Bioinformatics folder.
###########################################################
In-depth description of the files:
1) CNNs_Bioinformatics
This folder contains two subfolders, one for the
creation of the classification models and one for
their analysis and evaluation.
NOTE: In order to execute these scripts, the
MSClassifyLib needs to be included in MATLAB's
search path.
a) Training:
The contained files compute the models
for LDA with peak picking through their
Mann-Whitney-Wilcoxon statistic as well
as neural networks with the residual
network architecture and IsotopeNet.
NOTE: The file paths for DataTaskADSQ.mat
and DataTaskLP.mat need to be adjusted.
b) Evaluation:
analyzeCNNmodel.m creates various statistics
of the computed classification models.
evaluationBoxplot.m compares the various
models using box plots.
2) MSClassifyLib
This folder contains several subfolders, which
comprise the MSClassifyLib.
a) Core
This folder contains most of the required
definitions of the employed data types.
b) Classification
The employed classifiers are found here,
including the definitions of the used
neural network architectures.
c) ClassificationValidation
This folder contains classes and methods
with which a classification pipeline may
be defined.
d) FeatureExtraction
This folder contains classes and methods
for feature extraction, in particular
for computing the Mann-Whitney-Wilcoxon
statistic. Furthermore, 'dummy' feature
extraction methods are defined, which are
required for the classification pipeline.
e) Preprocessing
Several preprocessing methods. Here, only
the ones needed for the classification
pipeline are provided.
f) Helpers
Some small additional classes and methods.
##########################################################
ReadMe for the code of the article "Deep Learning for
Tumor Classification in Imaging Mass Spectrometry" by
Jens Behrmann, Christian Etmann, Tobias Boskamp,
Rita Casadonte, Jörg Kriegsmann, Peter Maass.
Should any questions arise, we would be happy to help you:
jensb@math.uni-bremen.de
cetmann@math.uni-bremen.de
The data is available at:
https://seafile.zfn.uni-bremen.de/d/85c915784e/
##########################################################
The provided files are structured in the following way:
1) CNNs_Bioinformatics
This is where executable MATLAB scripts are
stored which reproduce the results found in the
article.
2) MSClassifyLib
This folder contains parts of a larger library
developed at the Center for Industrial Mathematics,
Bremen. It provides routines for the analysis of mass
spectrometry data and is required for the execution
of the scripts in the CNNs_Bioinformatics folder.
###########################################################
In-depth description of the files:
1) CNNs_Bioinformatics
This folder contains two subfolders, one for the
creation of the classification models and one for
their analysis and evaluation.
NOTE: In order to execute these scripts, the
MSClassifyLib needs to be included in MATLAB's
search path.
a) Training:
The contained files compute the models
for LDA with peak picking through their
Mann-Whitney-Wilcoxon statistic as well
as neural networks with the residual
network architecture and IsotopeNet.
NOTE: The file paths for DataTaskADSQ.mat
and DataTaskLP.mat need to be adjusted.
b) Evaluation:
analyzeCNNmodel.m creates various statistics
of the computed classification models.
evaluationBoxplot.m compares the various
models using box plots.
2) MSClassifyLib
This folder contains several subfolders, which
comprise the MSClassifyLib.
a) Core
This folder contains most of the required
definitions of the employed data types.
b) Classification
The employed classifiers are found here,
including the definitions of the used
neural network architectures.
c) ClassificationValidation
This folder contains classes and methods
with which a classification pipeline may
be defined.
d) FeatureExtraction
This folder contains classes and methods
for feature extraction, in particular
for computing the Mann-Whitney-Wilcoxon
statistic. Furthermore, 'dummy' feature
extraction methods are defined, which are
required for the classification pipeline.
e) Preprocessing
Several preprocessing methods. Here, only
the ones needed for the classification
pipeline are provided.
f) Helpers
Some small additional classes and methods.
Markdown is supported
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