About

Background

One of the largest challenges in obtaining X-ray diffraction data from biological samples is growing large, high quality crystals.

Currently, there is no way to reliably predict successful crystallization conditions based on protein sequence alone and so high-throughput approaches are very appealing. High-throughput crystallization screens test a large chemical space using hundreds of different crystallization cocktails at the nano-drop scale. Successful conditions can then be scaled up and optimized to grow larger crystals.

The Crystallization Center at the Hauptman-Woodward Medical Research Institute provides this high-throughput screening service to users; offering 1536 condition screens for both soluble and membrane protein samples. Each plate is imaged over a period of two months in using both visible light microscopy and UV-TPEF photography.

This high-throughput produces a large volume of images that must be sorted through in order to pick out the best condition; a task that can be very tedious and repetitious.

In 2019 Bruno et al published Classification of crystallization outcomes using deep convolutional neural networks which included a CNN model that could accurately classify crystallization screening images, opening the door to automating this process. The MARCO model has been used in large scale projects such as the xtution database but has not been utilized in a average-user oriented graphical program.

Polo is therefore designed to incorporate the benefits of the MARCO model and integrate the functionality of established crystallization image labeling software such as MacroscopeJ to create a GUI targeted for HWI and other high-throughput crystallization screening users that incorporates all the tools needed to go from raw crystallization images to designing optimization screens without the need to install any dependencies.

Features

Automatic Image Classification

Using the MARCO model Polo can cut down the time you spend looking through your crystallization images by identifying wells likely to contain a protein crystal. This can reduce the total number of images that need to be considered from thousands to hundreds.

Multiple Image View Modes

Polo allows you to view your crystallization images in a variety of ways that make it easier to identify true crystal hits. Images can be viewed individually or in grids of up to 96. If a sample has been imaged at multiple points in time it is easy to create timeline views that allow you to assess the effectiveness of a screening condition over time. Additionally, if your samples have been imaged with photographic technologies outside of visible light microscopy it is easy to compare these images to verify the presence of protein crystals.

Integrated FTP Browser

Polo includes an simple FTP browser that allows you to download image files from a remote server directly into Polo without then need to install other software such as FileZilla. Polo is also packaged with unrar for Windows and Mac.

Data Management

Your image classifications are easily saved and managed via the xtal file format. Xtal files are similar to mso files created by MacroscopeJ and encode your image classifications, MARCO classifications, cocktail formulation and other metadata. In addition, xtal files increase portability by encoding your screening images directly into the file along side your metadata. This allows your classifications to be easily shared with one file to anyone else with Polo on their computer.

Polo also has options to export your runs to csv files without encoding your screening images or to HTML reports for a more visual way to share your results with those who do not use Polo.

Open Source Code Base

Polo is written in Python and is licensed under the GNU 3.0 license. This allows for modification and use of any of the Polo source code. If you wish to change modify or extend any of Polo’s functionality you are free to do so.