Electron Microscopy (EM) is rapidly growing as a high-resolution technique for structural biology. However, the high price of the electron microscopes is causing the concentration of scientific equipment in a few labs whose access is granted through infrastructure projects, like Instruct. A typical session in one of these high-end microscopes can generate up to 2TB of data per day, which is analysed by the researcher in its home institution. A new data acquisition will be performed when another access slot is granted. In this way, data analysis is coming at bursts, with high computational demands when the data arrives and low activity between acquisitions. Not all laboratories can justify the cost of a computer cluster with the necessary computational capabilities (in terms of number of cores, RAM memory, disk, GPU, …) with this intermittent use. Computer clouds offer a solution to this problem. This computing power is paid only on demand, and when there is no need of a computer for the data analysis, the machine is simply switched off with no associated cost (maintenance, amortization, dedicated installations and personnel, …).
EM software can easily run in the cloud. In fact, Scipion Cloud has been developed as an Instruct Pilot Project, and it is now publically available in the Amazon Cloud. The proposed course will have two aims:
– The first one will be to show how to create, setup, use, stop and terminate cloud machines, including the appropriate EM software, so that users can easily adapt the computer power to their needs.
– The second aim of the course will be to work with the course participants over a typical image processing workflow for single particle image processing.
The course will make extensive use of the I2PC software integration framework Scipion, accessing the most widely used software suites in the field, such as EMAN, Relion, Frealign and XMIPP.
The course is open to investigators at all levels. We will work with examples data, but participants can bring their own data to look at if they wish.