Science moves on

A Rebuttal to a Cool Climate Paper:
[Via Dot Earth]

A new paper challenges a cool forecast by a prominent skeptic of dangerous global warming.

[More]

A pretty reasonable view of how science moves forward. What is interesting is that the first comment is simply full of the same denialist crap that is constantly refuted. Some people just are incapable of learning.

There is a nice comment about the paper at RealClimate.

Technorati Tags: Environment

Sequencing driving storage

by Kables
In the cloud, Next Gen DNA sequencing computes:
[Via Discovering Biology in a Digital World]

These days, DNA sequencing happens in one of three ways.

In the early days of DNA sequencing (like the 80′s), labs prepared their own samples, sequenced those samples, and analyzed their results. Some labs still do this.

Then, in the 90′s, genome centers came along. Genome centers are like giant factories that manufacture sequence data. They have buildings, dedicated staff, and professional bioinformaticians who write programs and work with other factory members to get the data entered, analyzed, and shipped out to the databases. (You can learn more about this and go on a virtual tour in this nice video from Washington University).

At the same time, Universities expanded their core service laboratories and these labs began offering a greater number of sequencing services. Today, much the non-genome center sequencing happens in core labs. Scientists obtain samples and send them to the core labs. The core lab staff prepare the samples, carry out the sequencing reactions, and deliver data to their customers.

This system worked fine until the Next Generation DNA sequencing (NGS) instruments came along.

This January issue of Nature Biotechnology has two articles (1, 2) that address the role cloud computing can play in helping smaller laboratories cope with the large volumes of data produced by NGS.
[More]

The huge amount of data that can be generated using next generation sequencing techniques will pose some unique challenges for researchers. Using cloud computing has some significant advantages but also some unique problems.

While security is often a concern, I expect that market forces will provide some useful solutions. To me some of the worries are more fundamental when one cedes control of the data to someone else.

As seen just a few months ago, technical problems with the cloud servers can result in the loss of data. Researchers are putting a lot of trust that the owners of the servers do the proper backups and manage the facility properly.

Also, there could be concerns about the future of the servers that hold the data as companies change ownership. Processes can change with new ownership. Assuming there is a new owner. Bankruptcy is a worry when someone else runs the servers.

In fact, no one really has an idea of how long the data will need to be stored and available. The expectation is that there is no time limit for how long digital data can be stored but with changing technology and changing business cultures, this is uncertain.

Of course, a big problem is that the pipes of the Internet are not really designed to move terabytes of data around. As stated in one of the articles Sandra mentions, it can be easier putting the data on a hard drive and FedExing it than to use the Internet.

Storing data in the cloud will be an increasing part of scientific research; it has too many advantages. So I expect most of these worries to be dealt with. One thing for sure, storage manufacturers will be doing quite well over the next few years.

Technorati Tags: Bioinformatics, Science