Slashdot Mirror
Sequencing a Human Genome In a Week
blackbearnh writes "The Human Genome Project took 13 years to sequence a single human's genetic information in full. At Washington University's Genome Center, they can now do one in a week. But when you're generating that much data, just keeping track of it can become a major challenge. David Dooling is in charge of managing the massive output of the Center's herd of gene sequencing machines, and making it available to researchers inside the Center and around the world. He'll be talking about his work at OSCON, and gave O'Reilly Radar a sense of where the state of the art in genome sequencing is heading. 'Now we can run these instruments. We can generate a lot of data. We can align it to the human reference. We can detect the variance. We can determine which variance exists in one genome versus another genome. Those variances that are cancerous, specific to the cancer genome, we can annotate those and say these are in genes. ... Now the difficulty is following up on all of those and figuring out what they mean for the cancer. ... We know that they exist in the cancer genome, but which ones are drivers and which ones are passengers? ... [F]inding which ones are actually causative is becoming more and more the challenge now.'"
Functions that don't do anything, no comments, worst piece of code ever!
I say we fork and refactor the entire project.
Illumina will sequence your genome for $48,000.
http://scienceblogs.com/geneticfuture/2009/06/illumina_launches_personal_gen.php
Details.
How we know is more important than what we know.
Suppose they sequence a specific human's genome. Now they do it again. Will the two sequences be the same?
"Eve of Destruction", it's not just for old hippies anymore...
Just store all that data as a chemical compound. Maybe a nucleic acid of some kind? Using two long polymers made of sugars and phosphates? I bet the whole thing could be squeezed into something smaller than the head of a pin!
We pissed away $3 billion dollars and 13 years of time, when we could have waited a few more years and got it done in a week, and much, much cheaper. What a waste of time and money that was....
I know I'm being trolled, but you're an idiot. It's pretty obvious that the ability to sequence the genome in a week could only result from techniques developed and information gathered in the original Human Genome project.
I had but a simple dream, to destroy all humans.
It wasn't the computing power that was the holdup, it was the sequencing throughput. Also, as noted in the article, they can do it in a week now partially because they have the completed human genome to use as a template to match things up against. As I analogized in the interview, it's like the difference between putting together a jigsaw puzzle with the cover image available, and doing one without.
What's funny is that there is actually people who think like that. Apparently if we just sit around and wait, things will get better. I call this the dark side of the "invisible hand" of the market.. because it is invisible, people forget how it comes about. In order to get improvement in technology you need a market for that technology. And, typically, you need some loss-leader to create the market in the first place. Government funding serves this purpose well.
How we know is more important than what we know.
Data handling and analysis is becoming a big problem for biologists generally. Techniques like microarray (or exon array) analysis can tell you how strongly a set of genes (tens of thousands, with hundreds of thousands of splice variants) are being expressed under given conditions. But actually handling this data is a nightmare, especially as a lot of biologists ended up there because they love science but aren't great at maths. Given a list of thousands of genes, teasing out the statistically significantly different genes from the noise is only the first step. Then you have to decide what's biologically important (e.g. what's the prime mover and what's just a side-effect), and then you have a list of genes which might have known functions but more likely have just a name or even a tag like "hypothetical ORF #3261", for genes that are predicted by analysis of the genome but have never been proved to actually be expressed. After this, there's the further complication that these techniques only tell you what's going on at the DNA or RNA level. The vast majority of genes only have effects when translated into protein and, perhaps, further modified, meaning that you cant's be sure that the levels you're detecting by the sequencing (DNA level) or expression analysis chips (RNA level) actually reflects what's going on in the cell.
One of the big problems studying expression patterns in cancer specifically is the paucity of samples. The genetic differences between individuals (and tissues within individuals) means there's a lot of noise underlying the "signal" of the putative cancer signatures. This is especially true because there are usually several genetic pathways that a given tissue can take to becoming cancerous: you might only need mutations in a small subset of a long list of genes, which is difficult to spot by sheer data mining. While cancer is very common, each type of cancer is much less so; therefore the paucity of available samples of a given cancer type in a given stage makes reaching statistical significance very difficult. There are some huge projects underway at the moment to collate all cancer labs' samples for meta-analysis, dramatically increasing the statistical power of the studies. A good example of this is the Pancreas Expression Database, which some pacreatic cancer researchers are getting very excited about.
Illumina's Solexa sequencing produces around 7 TB of data per genome sequencing. Its a feat just to move the data around, let alone analyze it. Its amazing how far sequencing technology has come, but how little our knowledge of biology as a whole has advanced. 'The Cancer Genome' does not exist. No tumor is the same and in cancer, especially solid tumors, no two cells are the same. Sequencing a gamish of cells from a tumor only gives you the average which may or may not give any pertinent information about the tumor. Vogelstein's group has shown this quite convincingly but hardly anyone truly looks at what the data really says.
Functions that don't do anything, no comments, worst piece of code ever!
Most of it doesn't code proteins or any of the other things that have been reverse-engineered so far. How do you know it's NOT comments?
(And if terrestrial life was engineered and it IS comments, do they qualify as "holy writ"?)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
You have to be very careful about what findings at different levels actually mean, and how the various levels correlate.
For example when looking at duplications/expansions in cancer, an expansion of a locus results in about a 50% correlation between DNA level change and expression level chane. Protein and gene expression levels correlate 50 to 60% of the time (or less depending on who's data you look at). So therefore, being gracious and assuming a 60% correlation at the two levels you are already below a 40% correlation. Add in post translation modifications, sub-cellular localization and the requirement for other players within a functional pathway to exhibit a specific behavior and what you have is a tangled mess that you can spin almost any story about a favorite gene. But does it have meaning for diagnosis and treatment? I'd definitely hedge my bet.
Its that big of a mess and that isn't even considering the vast population heterogeneity of each tumor.
A good example of this is the Pancreas Expression Database, which some pacreatic cancer researchers are getting very excited about.
Kim Jong-il will be ecstatic to hear that. Dear Leader can't very well put the Grim Reaper into political prison....
Four bases and not much in between.
The human genome is approximately 3.4 billion base pairs long. There are four bases, so this would correspond to 2 bits of information per base. 2 * 3,400,000,000 /8 /1024 /1024 = 810.6 MiB of data per sequence. That doesn't seem like it'd be too difficult. With a little compression it'd fit on a CD. Now, I suppose each section is sequenced multiple times and you'd want some parity, but it still seems like something that'd easily fit on a DVD (especially if alternate sequences are all diff'd from the first). Perhaps throw in another disc for pre-computed analysis results and that ought to be it.
So, what's going on here? Are the file formats used to store this data *that* bloated? Or are they trying to include structural information beyond sequence? What am I missing that makes this an unwieldy amount of data?
(I have to laugh at how Vista is apparently 20 times more complex than the people that use it...)
Well, EENterestingly, that's pretty much what people are saying when they complain that early paleontologists ruined priceless artifacts.
You learn as you go, like when you're learning to play Ghosts 'n' Goblins and you keep getting killed by the red gargoyle, but then you eventually learn that you have to jump away from him as he swoops and fire frantically towards him. I know other people have made similar responses, but I only understand things in terms of analogies. Particularly ones related to throwing lances at gargoyles.
--
The vast majority of genes only have effects when translated into protein
That depends on your definition. If you define a gene as "stretch of DNA that is translated into protein," which until fairly recently was the going definition, then of course your statement is tautologically true (replacing "the vast majority of" with "all.") But if you define it as "a stretch of DNA that does something biologically interesting," then it's no longer at all clear. Given the number of regulatory elements not directly associated with genes, sections of DNA that code for RNAzymes, etc., it may well be that the majority of "genes" are not protein-coding at all. Going back to the Mendelian definition of a gene as a unit of inheritance, this looks more and more likely.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
a whole human genome will fit on a CD.
if you just transmit the diffs from the generic human you could put it in an e-mail
Some drink at the fountain of knowledge. Others just gargle.
I suppose it's worth noting that the intermediate (raw) data sets can get pretty large. they are actually getting larger as the trend goes towards shorter less informative "reads" that require more of them to recover the connective information and to recover from errors and duplications. However that's a tend that has a stopping point. While more reads is better at some point there is almost no added value from more reads. So at that point that's the maximum amount of data you need to collect. it's won't increase ever. meanwhile hard drive and network speeds will go up factors of ten.
thus the storage issues here are well tolerated at present and soon will become trivial.
Some drink at the fountain of knowledge. Others just gargle.
This actually suggests that perhaps we should start transmitting into space or on space crafts the genome of all the genes ever sequence, even the ones hauled out of the ocean that we don't know what organism they belong too. you send that, plus the molecular composition of DNA, and the molecular structure of the ribosome and T-rna
while there's more to a cell than just that, it's well known that in virto you can get transciption of the DNA from just that. It won't be too long I suspect before you could come up with some way to bootstrap a primordial cell out of those expressed proteins. Once you have such a cell, bootstrapping to higher level organisms is not such a long leap.
You would be effectively preserving an approximation of the earth's ecosystem. maybe someone will find it.
Some drink at the fountain of knowledge. Others just gargle.
A single run on a Solexa next gen sequencer can generate over 200GB of data and half a million files. And that is for 8 samples only. You get into the terabyte range very quickly.
That's why data is delivered on hard drives.
What does "sequence a genome" actually mean. The name "sequence" suggests that it has something to do with the "order" of something. Your post makes it sound like sequencing is something done before the computer gets ahold of the data. Can you explain for us genetics laypersons what the heck "sequencing" is? Tnx.
Very simply: Your DNA is stored in chromosomes. Each chromosome contains DNA in tight bundles with lots of weird secondary and tertiary structure. Suppose that you took all the chromosomes from one of your cells -- i.e., your genome -- unwound the DNA into long threads, and laid those threads out. You'd then have the chemical equivalent of strings of characters, e.g. ACGTGCATT ..., one for each chromosome, where each character represents a particular base. (I'm not going to get into the biochemistry, and anyway, there are probably people here who can explain it better than I can -- I'm a bioinformaticist, but mainly a numbers guy.) This ordered set of strings of characters is what's known as "the sequence," and "to sequence" a genome is to obtain that set.
Unfortunately, the actual sequencing process is a hell of a lot more complicated than what I just described, and considerable computational power is required at all stages of the process. But really the number crunching isn't the bottleneck, it's the biochemistry. And that's been improving rapidly, so now we have the ability to do the "wet-lab" work necessary to get an entire human (or any other organism) genome sequence a lot faster than we used to.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
Next gen sequencing eats up huge amounts of space. Every run on our Illumina Genome Analyzer II machine takes up 4 terabytes of intermediate data, most of which comes from the something like 100,000+ 20 Mb bitmap picture files taken from the flowcells. All that much data is an ass load of work to process. Just today I got a little lazy with my Perl programming and let the program go unsupervised...and it ate up 32 gb of ram and froze up the server. Took redhat 3 full hours to decide it had enough of the swapping and kill the process.
For people not familiar with current generation sequencing machines, they can scan between 30-80 bp reads and use alignment programs to match up the reads to species databases. The reaction/imaging takes 2 days, prep takes about a week, processing images takes another 2 days, alignment takes about 4. The Illumina machine achieves higher throughput than the ABI ones but gives shorter reads; we get about 4 billion nt per run if we do everything right. Keep in mind though, that 4 billion that they mention in the summary is misleading: the read cover distribution is not uniform (ie you do not cover every nucleotide of the human's 3 billion nt genome). To ensure 95%+ coverage, you'd have to use 20-40 runs on the Illumina machine...in other words, about 6-10 months of non-stop work to get a reasonable degree of coverage over the entire human genome (at which point you can use programs to "assemble" the reads in a contiguous genome). WashU is very wealthy so they have quite a few of these machines available to work at any given time.
the main problem these days is that processing all that much data requires a huge amount of computer knowhow (writing software, algorithms, installing software, using other people's poorly documented programs), and a good understanding of statistics and algorithms, especially when it comes to efficiency. Another problem they never mention are artifacts from the chemical protocol; just the other day we found a very unusual anomaly that indicated the first 1/3 of all our reads was absolutely crap (usually only the last few bases are unreliable); turned out our slight modification of the Illumina protocol to tailor it to studying epigenomic effects had quite large effects of the sequencing reactions later on. Even for good reads, a lot of the bases can be suspect so you have to do a huge amount of averaging, filtering, and statistical analysis to make sure your results/graphs are accurate.
What's funny is that there is actually people who think like that. Apparently if we just sit around and wait, things will get better. I call this the dark side of the "invisible hand" of the market.. because it is invisible, people forget how it comes about. In order to get improvement in technology you need a market for that technology. And, typically, you need some loss-leader to create the market in the first place. Government funding serves this purpose well.
The sad thing is that this seems to be pretty much par for the course. If only we wait just a little while and skip all those annoying intermediate steps, we will soon have fantastically good rockets / fusion reactors / whatever else without having to pay anything...
If it has taken 13 years until recently to properly sequence the genome of a single human, how has it been possible to do DNA "fingerprinting" e.g. for crime investigations? Is the actual sequencing not required there?
yes, let's give those aliens something to experiment on, so they can figure out what bugs to send our way to exterminate us :)
MP3 Search Engine
Well, how about pollution, processed food, and all that trash being the main reason we get cancer?
Cancer was not even a known disease, a century ago, because nobody had it. (And if people get cancer now, way before the average age of death a century ago, then it can't be that it is because we now get older.)
But I guess there is no money in that. Right?
Any sufficiently advanced intelligence is indistinguishable from stupidity.
Fingerprinting doesn't rely on DNA sequencing, but does rely on the DNA sequence being different between people. Everyone's DNA contains subtle differences (particularly in the non-coding DNA regions). These differences can be exploited by various laboratory techniques to produce small pieces of DNA which will be of different sizes because of these differences. When these fragments of DNA are run down a suitable gel (usually agarose, a substance derived from seaweed) under an electric current the fragments will separate by size. The pattern of fragments formed will be unique for each individual.
Several fingerprinting techniques rely on what most programmers would best recognise as regular expression matching. For example there are enzymes in biology which will recognise certain DNA sequences but not others, and will cut the DNA in two where ever this sequence is matched. (in perl:
is the equivalent of what an enzyme called EcoRI does). Not everyone will have the same numbers of this sequence in their DNA, and nor will they be in the same place, thus the number and size of fragments will differ. By using a suitable range of such enzymes you can generate a pattern of DNA fragments which is sufficiently unique as to identify a single person amongst a population of several billion.
for more information take a look at DNA Profiling on wikipedia
...used against me for anything without violating the DMCA. The act of decoding it by some forensics lab paternity test or future insurance company medical cost profile would become unlawful and I'm sure the RIAA would help me with the cost of prosecuting the lawsuit.
The problem with quotes on the internet, is that nobody bothers to check their veracity. -- Abraham Lincoln
Check the vending machines !
Squirrel!
First, kinds of cancers were known to exist a century ago. Tumors and growths were not unheard of. Most childhood cancers killed quickly and were undiagnosed as specific disease other than "wasting away". When the average lifespan was 30-40 years, a great many other cancers were not present because people didn't live long enough to die from them.
As we cure "other" diseases, cancers become more likely causes of death. Cells fail to divide perfectly, some may go cancerous others simply don't produce as healthy a replacement specialized cell. Your arteries harden, muscles don't repair as well, other tissues don't work as well (you get weaker, more wrinkled, easier to fall ill). Eventually either something fails that can't be repaired or enough cells go cancerous. Until we either figure out how to replace the body (seems unlikely as the brain and body are more tied together than sf movies like to present) or we figure out how to make cells repair/refresh themselves without shortening their telomeres -- I have no idea how likely that actually is.
The problem with quotes on the internet, is that nobody bothers to check their veracity. -- Abraham Lincoln
Very good points. I think I've been using a very sloppy definition of gene, just a vague idea that it's only DNA>RNA>protein>action or DNA>RNA>action. I've never really got deeply into thinking about regulatory elements, etc. It's compounded by the fact that, while I'm interested in cancer, most of my actual work is with a DNA-based virus that only produces a very few non-translated RNAs that we're aware of. I have a tough time convincing some people that even those are biologically relevant.
I sometimes think that RNAs and various epigenetic factors (I'm including DNA secondary and tertiary structures here) fall into the same trap as a lot of post-translational protein modifications: They're hard to study so not much is written or understood about them, so most non-specialists basically ignore them and decide they can't be too important. It's changing now as techniques evolve to do the experiments, but I'm still shocked how often I see someone basically say "well we don't understand this so we'll assume it's not affecting our system".
I'm curious how you figure 200GB of data. A solexa 1G only produces tens of millions reads per run, each read being about 36 bases.
Some drink at the fountain of knowledge. Others just gargle.
There are raw intensity files which are used for base calling. The output of the base calling is used to generated alignments, quality, etc. You don't just get the short reads at the end. Most people are just going to use the short reads but that still can be 30G of data for a run.
Splice too much of that bad, useless, convoluted code into a "new" human and we might end up with a G-Gnome or GNOME (Gratuitous, Nacent, Ogreous, Mechanised Entity). Call it... "G-UNIT", and give it a uniform and a mission. Or, give it a script and a part and call it Smeegul/Smigel...)
Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
While a single human genome is a lot of information, storing thousands shouldn't add much requirements, one can simply store a diff from the first.
[]'s Victor Bogado da Silva Lins
^[:wq
Illumina will sequence your genome for $48,000.
http://scienceblogs.com/geneticfuture/2009/06/illumina_launches_personal_gen.php
Details.
Helluva lots of details. After wasting some perfectly useful clicks all I could come up with was:
Illumina's technology is extremely well-established, and serves as the backbone for most large-scale genome sequencing projects currently underway (including the majority of the samples sequenced as part of the 1000 Genomes Project); that gives it an edge over the more experimental technology employed by competing sequence provider Complete Genomics.
That's so much details that I really want my clicks back. You should be ashamed of yourself, Sir.
This actually suggests that perhaps we should start transmitting into space
In an infinite universe, our DNA already exists out there. Its just a question of how far. Philosophically, there is nothing to be gained by pumping it out to our possible neighbours. It could only be used against us.
The raw images from the device alone can take up this much space. 8 lanes, 300 imaging regions (tiles) per lane. Each imaged 4 times (one for each base/channel). A typical run is 37 cycles (base pairs), paired end runs (now typical) double this so:
8*300*4*37*2 =710400
On a GA1 those files are 2mb each, giving you around a terabyte and a half of of primary data to process. Image analysis takes place processing those files in to "intensity files". Those are further processed in to corrected intensities, then basecalls. Each of these steps produces a similar number of files. Some details of the process here: http://sgenomics.org/mediawiki/upload/8/80/Pipeline.pdf
Those numbers are for a GA1, the current version of the instrument has less imaging regions (100). However cycle length has increased (typically now 75+ bp).
As a side note all the tools used are "shared source" and not available under an open source license. There is a project called Swift which is an open source tool to do this: http://bioinformatics.oxfordjournals.org/cgi/content/abstract/btp383