In the era of Clouds we still need the Bricks

A recent article in New York Times underlines once again the indredible rapid excalation in genomic data production boosted by NGS technology and the set-up of huge sequencing center, like the famous BGI (a monster with 162 HiSeq2000 sequencher capable of producing about 2,000 genomes/day).

So now it seems that the real bottleneck in genomics is no more the data production but its storage, analysis and sharing. The analysis step open a large new business for hi-tech and bioinformatic companies and rapid progresses are expected. However I think that even with super-fast computers and super-efficient analysis pipelines we can not multiply brains the researchers' heads and the problem to found the biological meaning of all these data are the challenge of today and tomorrow research.

But the real challenge come with the sharing part...In these day with everybody speaking about cloud technologies, rapid data transfers and the almost futility of local storage, we are in fact facing the hard problem of delivering the genomic data from the service provider (say the BGI in china) to the researcher who have to analyze the results (say me in my little lab in Italy). As already pointed out in a review on Nature, the Tb scale data easily reached by genomic projects require to get back to Bricks, also known as the old, concrete hard drives. The experts at BGI agreed that the most rapid and effordable way to deliver the results is to save them in a hard-drive and send it well packed with FedEx. In someway it is like we are suddenly back from the actual time of mobile smartphones to the times of hand written letters...Think to a large consortium, involving different research unit located in differetn countries, working on a large genomic project: the work will be consistently slowed down and the costs will raise since every unit in the consortium have to wait days for the results and analysis to be physically delivered  on a hard-drive...We are no more in the dreamy world of instant transfer. Of course after initial analysis you only have to move a small subset of the total data, but it still the fact that we are facing a situation in which moving the researchers to data could become cheaper than move the data to the labs and analyzing and moving the data would cost more than resequence.

So consider a barrow in your lab equipment: it's cheap and with NGS expanding there will be a lot of bricks to move!