Q&A page allows you to submit questions about our programs and exchange ideas with others.
Breast cancer histopathological images classification using a hybrid deep neural network
Integration of Multimodal Data for Breast Cancer Classification Using a Hybrid Deep Learning Method
updates: The “-f” mode will align tilt series according to the given angle file regardless of instrument imprecision (i.e., the “-f” assumes that the platform is quite stable and has little vibration, and would not refine the tilt angles). The default ( without “-f” ) mode will refine the tilt angles and produce refined tilt series that consistent with the transformation.
Related materials of “DLBI: deep learning guided Bayesian inference for structure reconstruction of super-resolution fluorescence microscopy.”
Related materials of “DLBI: deep learning guided Bayesian inference for structure reconstruction of super-resolution fluorescence microscopy.” DLBI includes three main parts, stochastic simulation, deep neural network and Bayesian inference. There we combine the main idea of SIMBA and DLBI and provide a script to do the analysis of all three parts.
There is a folder includes all files of DLBI and SIMBA, and you can just use the simba.sh to process the data. For further information, please read ReadMe.
Download: DLBI-master.zip (75 downloads)
A Hybrid Convolutional and Recurrent Deep Neural Network for Breast Cancer Histopathological Image Classification
Related materials of “Live-cell single molecule-guided Bayesian localization super-resolution microscopy”
Related materials of “Live-cell single molecule-guided Bayesian localization super-resolution microscopy”. These datasets includes two actin structure and one Endoplasmic reticulum structure.
These datasets are owned by Professor Pingyong Xu’s group, and any people who want to use these data should cite the paper “Live-cell single molecule-guided Bayesian localization super-resolution microscopy” and “Rational design of true monomeric and bright photoactivatable fluorescent proteins”. For further information, please contact firstname.lastname@example.org, email@example.com.
AuTom-dualx is a toolkit for fully automatic alignment of dual-axis tilt series with simultaneous reconstruction. It provides global consistent alignment with projection model with different complexity. The default alignment is based on the distortion correction along x, y and z direct. The users are recommended to use this model but a high-level distortion correction is also available by option “-w 2”. Here we give out the pre-compiled exe and test datasets as well as their anticipated results:
dualxmauto(source code): dualxmauto-warp-correction.zip (132 downloads)
volrec_mltm_ubuntu14 (164 downloads)
volrec_mltm(source code): volrec_mtlm.tar.gz (125 downloads) This code has been compiled and run with open MPI 1.8.3 version successfully
Here are the examples demonstrated in the manuscript “Dualxmauto: a toolkit for fully automatic alignment of dual-axis tilt series with simultaneous reconstruction”.To run the examples, module dualxmauto and volrec_mltm is needed.If you extract the tar of testdata, a cmd.txt can be found. Following the cmd.txt you can get the result as in the manuscript or _Anticipated_result.
tools: tools.tar.gz (126 downloads)
markerauto 1.6.3 version
If the reader run as “markerauto -i V4B_G1_Tilt2.mrc -a V4B_G1_Tilt2.rawtlt -d -1 -o V4B_G1_Tilt2.xf -n V4B_G1_Tilt2.tlt”, markerauto1.6+ will use the old fasion to process the data. If the reader run as “markerauto -i V4B_G1_Tilt2.mrc -a V4B_G1_Tilt2.rawtlt -d -1 -o V4B_G1_Tilt2.xf -n V4B_G1_Tilt2.tlt -t”, markerauto1.6+ will use the Gaussian mixture model based fiducial marker tracking to process the data, which is much fater than the old fashion. The “-t” mode will use GMM-based model and the old fashion will use RANSAC-based model, users can make a execution comparion between these two model. However, because GMM-based model is in its first stage, it is still not so robust. Users are advised that use RANSAC-based model as default and use GMM-model for large field dataset.
program: markerauto_Centos6.5.tar.gz (325 downloads) , markerauto_RedHatEnterprise6.4.tar.gz (306 downloads) , markerauto_Ubuntu14.tar.gz (143 downloads) , markerauto_Ubuntu16.04.tar.gz (320 downloads)
Automatic Tomography (Au-Tom) is for automatic reconstruction of electron tomography (ET), which covered the pre-processing, alignment and reconstruction of electron tomography. In our package, fiducial marker-based datasets and maker-free datasets are done with totally different subprocess. The presented package has the following characteristics: accurate alignment modules for datasets that contain substantial biological structures but free of fiducial markers;fully automatic alignment modules for datasets that have fiducial markers embedded in; a wide coverage of reconstruction methods with a new iterative reconstruction method that recovers “missing wedge” based on compressed-sensing theory; multi-platform acceleration solutions that support faster iterative algebraic reconstruction. Currently, the markerbasd alignment and reconstruction of AuTom is the mostly well challenged module, while the markerfree alignment still has the limitations of data features. Autom has been built under Red Hat Enterprise 6.4, Cenots 6.5,Ubuntu 14.04 and Ubuntu 16.04. Other systems may not be supported well.
Read the installation guide to learn how to install Autom. You need to 1.download software package of AuTom. 2.Install Library dependencies listed in installation guide: CUDA8, mpich2-1.4.1, gnuplot-5.0.5, libgd-2.1.0 and other dependencies. 3.Click Autom or type “./Autom” in terminal to run the Program. The tutorial and videos (markerbased_video, markerfree_video) show you how to use Autom step by step. Apart from the user iterface, you can also access the modules in the terminal. Please read the user-guides inside the doc package.
7.A few bug fixes.
If you have any problems or feedback, please submit your question or e-mail us. We really appreciate receiving your advice as soon as possible. E-mail: firstname.lastname@example.org, email@example.com.
The publication describing AuTom is “Han R, Wan X, Wang Z, et al. AuTom: a novel automatic platform for electron tomography reconstruction[J]. Journal of Structural Biology, 2017, 199(3):196”. Please cite this article in your essay if you use AuTom.
ICONv1.6.4 & ICONMICv1.0.0
Iterative Compressed-sensing Optimized Non-uniform fast Fourier transform reconstruction (ICON) based on the theory of compressed-sensing and the assumption of sparsity of biological specimen. ICON-GPU_v1.2.8_CentOS64.tar (70 downloads) ICON-GPU_v1.2.8_Ubuntu64.tar (47 downloads) ICON-GPU-v1.2.8_UserGuide.pdf (47 downloads) ICON-MICv1.0.zip (155 downloads)
ET-SPEC(formerly known as ETphantom)
ET-SPEC is a virtual electron microscopy molecular modeling simulation and 3D reconstruction system based on serial block face scanning electron microscope image(SBEM).The package contains installation instruction and a sample date. ET-SPEC.zip (220 downloads) The source code of ET-SPEC is available. sourcecode.zip (143 downloads) A Tutorial on getting started with ET-SPEC. Make sure you have installed softwareforET-SPEC.zip (197 downloads) . Dual-axis nowarp,marker,projection,tracking result generated by ET-SPEC. test_mean_nowarp.tar.gz (471 downloads) test_mean_nowarp_rot90.tar.gz (109 downloads) Dual-axis warp,marker,projection,tracking result generated by ET-SPEC. test_mean_warp.tar.gz (426 downloads) test_mean_warp_rot90.tar.gz (101 downloads)
markerauto 1.5.3 version
markerauto 1.5.0 version
updates: 1. The initial diameter value can be automatic detected, if the user set the initial input to -1, for example:markerauto -i BBa.st -a BBa.rawtlt -n BBa_new.tlt -o BBa_fin.xf -d -1 2. fixed the bug when proceeding fiducial markers with large diameter. 3. fixed the bug for minimum number of fiducial markers (minimum number: six fiducial markers). markerauto (213 downloads)
markerauto alpha version
Automatic marker_based alignment module.Material data contains the exe and two small test data (1024×1024). Test_106 contains a test data (with 2048×2048 size). test_106.tar.gz (626 downloads) material_data_1.tar.gz (246 downloads)
atomalign alpha version
dualxmauto-warp-correction.zip (132 downloads) Autom.pdf AuTom: A novel automatic platform for electron tomography reconstruction (48 downloads) AuTom: A novel automatic platform for electron tomography reconstruction (48 downloads)