Volumetric super-resolution imaging by serial ultrasectioning and stochastic optical reconstruction microscopy in mouse neural tissue.

Here, we present a protocol for collecting large-volume, four-color, single-molecule localization imaging data from neural tissue. We have applied this technique to map the location and identities of chemical synapses across whole cells in mouse retinae. Our sample preparation approach improves 3D STORM image quality by reducing tissue scattering, photobleaching, and optical distortions associated with deep imaging. This approach can be extended for use on other tissue types enabling life scientists to perform volumetric super-resolution imaging in diverse biological models. For complete details on the use and execution of this protocol, please refer to Sigal et al. (2015).

Trainable high resolution melt curve machine learning classifier for large-scale reliable genotyping of sequence variants.

High resolution melt (HRM) is gaining considerable popularity as a simple and robust method for genotyping sequence variants. However, accurate genotyping of an unknown sample for which a large number of possible variants may exist will require an automated HRM curve identification method capable of comparing unknowns against a large cohort of known sequence variants. Herein, we describe a new method for automated HRM curve classification based on machine learning methods and learned tolerance for reaction condition deviations. We tested this method in silico through multiple cross-validations using curves generated from 9 different simulated experimental conditions to classify 92 known serotypes of Streptococcus pneumoniae and demonstrated over 99% accuracy with 8 training curves per serotype. In vitro verification of the algorithm was tested using sequence variants of a cancer-related gene and demonstrated 100% accuracy with 3 training curves per sequence variant. The machine learning algorithm enabled reliable, scalable, and automated HRM genotyping analysis with broad potential clinical and epidemiological applications.

MIPModDB: a central resource for the superfamily of major intrinsic proteins.

The channel proteins belonging to the major intrinsic proteins (MIP) superfamily are diverse and are found in all forms of life. Water-transporting aquaporin and glycerol-specific aquaglyceroporin are the prototype members of the MIP superfamily. MIPs have also been shown to transport other neutral molecules and gases across the membrane. They have internal homology and possess conserved sequence motifs. By analyzing a large number of publicly available genome sequences, we have identified more than 1000 MIPs from diverse organisms. We have developed a database MIPModDB which will be a unified resource for all MIPs. For each MIP entry, this database contains information about the source, gene structure, sequence features, substitutions in the conserved NPA motifs, structural model, the residues forming the selectivity filter and channel radius profile. For selected set of MIPs, it is possible to derive structure-based sequence alignment and evolutionary relationship. Sequences and structures of selected MIPs can be downloaded from MIPModDB database which is freely available at http://bioinfo.iitk.ac.in/MIPModDB.

PDBalert: automatic, recurrent remote homology tracking and protein structure prediction.

BACKGROUND: During the last years, methods for remote homology detection have grown more and more sensitive and reliable. Automatic structure prediction servers relying on these methods can generate useful 3D models even below 20% sequence identity between the protein of interest and the known structure (template). When no homologs can be found in the protein structure database (PDB), the user would need to rerun the same search at regular intervals in order to make timely use of a template once it becomes available. RESULTS: PDBalert is a web-based automatic system that sends an email alert as soon as a structure with homology to a protein in the user's watch list is released to the PDB database or appears among the sequences on hold. The mail contains links to the search results and to an automatically generated 3D homology model. The sequence search is performed with the same software as used by the very sensitive and reliable remote homology detection server HHpred, which is based on pairwise comparison of Hidden Markov models. CONCLUSION: PDBalert will accelerate the information flow from the PDB database to all those who can profit from the newly released protein structures for predicting the 3D structure or function of their proteins of interest.