i was interviewed by "wired" a few years ago and the editor called me to confirm the story and all quotes. i guess that wasn't a normal practice for them.
If it is so that this was your choice (unlike some people that do it just to get free education -- I can sympathise a bit with these guys), you shouldn't expect everyone to feel sorry for ya.
ummm, excuse me, but there is no draft anymore, so everyone who joins the military does so voluntarily and with full knowledge of the potential consequences (unless they are truly dumbfucks). and i have no sympathy whatsoever for the people who join the military just to get a free ride in terms of schooling. i mean, come on! it's the frelling military! what the hell do you think it's purpose is? to hand out free educations? i think not.
with the recent release of digital performer 3.1, i heard that MOTU has put all developers onto the osX version which i heard is in beta. 3.1 is supposedly the last os9.x version.
i can't wait until i can do all my music stuff in osX, but in addition to MOTU i have to wait for native-instruments to get going on reaktor for osX.
hopefully, the protools announcement will help others get their asses in gear.
not ludicrous at all. a mechanism of the evolution of new genes may be as follows:
gene gets duplicated sans its regulatory elements (so, it cannot be expressed under the same conditions as the gene it was duplicated from), but is not protected from mutations, so over time (and not necessarily a long time) it begins to accumulate mutations.
such things have been documented in the genome. they are called "pseudogenes." most current thought is that these are just junk, remnants, but i think that these may actually be the source of new genes--genes caught in the act of evolving.
then, in a time of stress, there is a mechanism that increases the expression of these genes. i think that this must be what the article is talking about. though, i've not read the article, so i cannot be sure.
and my opinions here are not uninformed--i do this stuff for a living.
has the guy never heard of os x? or does he not want to get new hardware? it seems to me that os x is the best of both worlds.
james
Re:Learning Bioinformatics
on
Biohackathon
·
· Score: 1
I mostly agree with what you say, except that in my particular case, i think that i fall in the second category... I know what the data I generated means, I understand a lot of their implications (but not all of them:( ), I have gotten accustomed to extracting the information I need from northern blots, cleavage essays and stuff like that...
well, that's great, then! the more and varied projects you get to work on, the greater your knowledge-based will become and the more connections you'll be able to make in the data. good luck!
james
Re:Bioinformatics as a student
on
Biohackathon
·
· Score: 1
1) Get a Ph.D and become a card-carrying member of the fraternity.
bah! can you tell i disagree?;) actually, do what you _want_ to do...i feel that a phd is not required...it's certainly not bad to have, of course!
james
Re:Learning Bioinformatics
on
Biohackathon
·
· Score: 2, Insightful
Altough I agree that having a "wet" lab experience will have an impact on your computational research, i also think that having biology knowledge isn't required. Projects are usually narrow enough that all you need to understand is your local view of the problem.
as i said, it depends on what you want to do and what you want to get out of it. anyone with a narrow view and some programming abilities can generate data. it's the interpretation of the data that gets the discoveries...
I'm a computer science student, and i am enrolled in a coop program. Having basically no knowledge in biology, I have been able, within 3 months of work, to learn everything i needed and to produce enough results/data to create 3+ M.Sc./Ph.D wet lab projects.
good for you (i really mean that!). but can you analyze the data you generated and make useful conclusions that further the understanding of that particular project/field?
what you describe falls into the type 1 category i mentioned...and don't get me wrong--it's a very important category. it's just not what excites me...i'm more about having a relevant biological question (what is the function of gene x? is gene y involved in disease z? etc.) and then figuring out what tools i need (and create them if they're not already extant) and what data sources i need to mine. i can pump out gigs of data, but, in my mind, it's all about what useful information one can glean from it, not about how much you can produce. (and i'm not saying that that is what you were saying)
the more projects you become involved in, the more your knowledge base grows...be sure to have the scientists you're working with explain the _entire_ project to you so that you can see the bigger picture of what they are trying to accomplish...and maybe you'll be able to bring a different way of thinking to the project...it's not all about 1's and 0's.;)
james
Re:Learning Bioinformatics
on
Biohackathon
·
· Score: 2, Interesting
what do in need to learn to become useful in
bioinformatics?
i guess it depends on what it is you really want to do in the field..._very_ basically, there are two areas in bioinformatics: 1. the programmer who creates (possibly enterprise-level) tools as directed by the needs of the scientists and, 2. the bioinfomatics researcher/scientist who also develops tools at need, but also analyzes the data and makes conclusions and uses those conclusions/interpretations to guide wet lab work. and then, the results from the wet lab work come back to the bioinfo scientist who then incorporates the data to refine their ideas or to develop new ones which then go back into the lab. it's a very nice positive feedback loop when it works.
i fall in the latter category, which i like to call "genome hacking." the programming focus is to get the data and process it rather than making a tool that looks pretty, is user friendly, etc.
what i have found most useful in this regard is an extensive background in molecular/cellular biology (i have ~10 years of wet lab experience interspersed with my bioinformatics work (i've been full time bioinfo since '95/'96)). since molecular/cellular biology data is inherently noisy, i find that experience actually working with it and interpreting it has a profound impact on how i do my computational research as not only do i know what the wet lab is capable of doing, but i am also able to analyze wet lab data and make informed decisions based upon it...many times, this noise i spoke of has a story to tell...and sometimes it does not. it is experience that allows one to make the differentiation.
as to the type 1 bioinfo type, i always think that it is a good idea to have a working knowledge of the type of data you're processing--not just the form the data take (ie this is a text file, this is an image, etc.), but rather "this is a DNA sequence that may have errors in it and i need to be aware of that and know the types of errors that can occur so that i can include provisions for that." of course, it's more complicated than that, but i think you get the idea. of course, the best way to learn this is by doing...reading some basic molecular biology texts wouldn't hurt either.;)
But when you're doing something like clustering, well, you know, I'd rather have a big manual, thanks.
ummm...the one page document contains the instructions on how to set up the macs for clustering. the 200 page document contains the instructions on how to do it for linux.
why, in god's name, would you want to follow instructions that are 200 times longer?! call me crazy, but i'd rather spend the extra time actually _using_ the clustered machines rather than simply setting them up.
"The 'reporter' tag defines a sequence of codons (the four amino acids that comprise DNA)"
sheeesh! can't they even get the basics right? a codon is a unit of
three nucleotides
that encode a single amino acid (there are three out of the 64 that do not code for an animo acid, rather, they code for the translation stop signals).
four nucleotides comprise DNA. there are 20 amino acids.
Describes a perl script that can create a virtual DNA and corresponding protein sequence with user-defined characteristics such as choice of localization (secreted, nuclear, transmembrane, etc.) and presence of particular protein domains. Resulting sequence would be considered real and coding by most prediction methods. This will allow you to file composition of matter patents on your very own custom genes.
Maybe I'm just being overly paranoid, but this really seems like something which shouldn't be messed with. I've heard people say that Cancer and AIDS/HIV are nature's way of dealing with overpopulation.
Sure, but if you or a loved one had either disease would you refuse treatment? The same thing could be said about polio, small pox, and any other disease. The human race got to where it's at now by overcoming these obstacles...we grow by defeating that which wants to kill us.
You take antibiotics for granted. In 10-20 years your kids will take gene therapy for granted...
Besides, I think that the human race itself will probably be its own population control--wars, famine, etc. I hope not, but...
The suggestion is that Celera have been doing only this, filing patents solely based on sequence homology.
Yes, this is true though they're using more than simple sequence homology when they have it. Plus, they have a big effort to obtain the actual physical entity: a full length cDNA clone (this can be a very difficult and very costly undertaking). But I think it's OK since they do have to back it up with "proof" within a year or so. I'm sure they'll let the vast majority of the applications lapse. Of course, I could be wrong.
It should also be noted that other companies can work on the same clone up until the time that the actual patent issues. Or so I've been lead to believe.
To discover the make-up of a protein, with a sequence of a previously unknown type, which blocks incoming self-destruct signals inside the cell, and then to prove beyond doubt that this is what the sequence does, as jrg's team did -- that seems to me a lot more impressive than just going through data automatically, and identifying everything that looks like something somebody else has seen before.
Funny you should mention the automatic data analysis because I do some of that now, having moved entirely out of the lab and fully into computational biology/bioinformatics. In this case, I feel there are merits to obtaining provisional applications based only on computational evidence. You don't simply have to rely on strict sequence homology, but you can use HMMs, motifs, and any number of other computatioanl methods to try to identify a putative function for a novel gene. Note that this is for the provisional application only and it must be backed up within a year with actual "proof" for the application to proceed. Also note that the above mentioned patent is a continuation of previous ones that we filed when we did not yet have a full length clone, but we did know a good amount about its biology...
There is simply way too much data for laboratories to keep up with--the only feasible method of sorting it out is to analyze it computationally to try to categorize novel into genes into existing families and thus have a narrower focus when it comes time to prove your prediction with actual cellular biology experiments. Helps to save time, effort, and money.
I checked out the first link and from what little I read, they have identified a novel sequence which is homologous to a known gene family that has some known functions (the prior art). Therefore, the utility of the discovery is clear. In addition to this they have provided tissue distribution and show that many of the tissues the gene is expressed in are associated with the particular disease states that the gene family is.
I'm a scientist, not a patent laywer, so gross oversimplifications follow: When one files a provisional patent application on a gene, one is not applying for a patent on the gene itself, but rather on a _use_ for that gene. For example, "gene_x for use in treating osteoporosis." You then usually have a year to prove your claims by providing data or at least showing that you are actively working on it rather than just sitting on it. If another company or academic institution finds that gene_x is actually involved in cancer, then they can pursue gene_x for use in treating cancer and you're SOL (shit outta luck).
The key is proving utility and this is why ESTs are unpatentable en masse. The patent office has set a very high threshold for proving utility when it comes to biotech patents; I've personally had to field questions from patent examiners and they can be brutal--for good reason: now that cDNA and sequencing technologies are so ubiquitous, pretty much anyone can generate sequence data, but not everyone can prove utility; proving utility requires time and money and, in most cases, innovation. A provisional application is as they say, "a stake in the ground" showing that you were here first--you identified this gene and you think it is involved in process X. If you can prove this, the patent may be issued to you.
I personally feel that the genome should be open source (and it will be), but I do feel that people who mine the data in clever ways and find interesting things should have legal protection for their discoveries. And let's face it: many drugs on the market are there due to the research of for-profit companies (they have bigger resources, harder timelines, etc. than academia) usually in collaboration with academia. If companies cannot profit on drugs they discover, then I wonder how much the rate of discovery of new drugs for treating diseases will decrease...
As for the whole Celera deal...I don't understand what people are getting so huffy about. I've spoken to Ventner and Celera and they've always planned to release the entire genome to the public, gratis. True, they aren't going to release it as soon as they get it (nor will they release all the work and analysis they've done on it), but the sequence itself will be released. Nothing they do with it will prevent academics from doing research...it may simply limit what other companies can do. It may be a moot point though since Celera at one point did get some DOE funding and this might make the United States Government a co-applicant of these patents. Co-applicants can make any patented claims public domain and the USG is likely to do this depending on how it all works out.
That's my 2 cents on a very complicated issue, james
Slashdot Mirror
This sounds more like a stealth lay-off than some sort of efficiency-promoting move.
Winn Schwartau
;P
isn't he on the jedi council?
i was interviewed by "wired" a few years ago and the editor called me to confirm the story and all quotes. i guess that wasn't a normal practice for them.
;)
i bet it will be now.
james
don't these astronomers read lovecraft?
should this not be called, "yuggoth"?
If it is so that this was your choice (unlike some people that do it just to get free education -- I can sympathise a bit with these guys), you shouldn't expect everyone to feel sorry for ya.
ummm, excuse me, but there is no draft anymore, so everyone who joins the military does so voluntarily and with full knowledge of the potential consequences (unless they are truly dumbfucks). and i have no sympathy whatsoever for the people who join the military just to get a free ride in terms of schooling. i mean, come on! it's the frelling military! what the hell do you think it's purpose is? to hand out free educations? i think not.
james
with the recent release of digital performer 3.1, i heard that MOTU has put all developers onto the osX version which i heard is in beta. 3.1 is supposedly the last os9.x version.
i can't wait until i can do all my music stuff in osX, but in addition to MOTU i have to wait for native-instruments to get going on reaktor for osX.
hopefully, the protools announcement will help others get their asses in gear.
james
This is ludicrous. Bottle up?
not ludicrous at all. a mechanism of the evolution of new genes may be as follows:
gene gets duplicated sans its regulatory elements (so, it cannot be expressed under the same conditions as the gene it was duplicated from), but is not protected from mutations, so over time (and not necessarily a long time) it begins to accumulate mutations.
such things have been documented in the genome. they are called "pseudogenes." most current thought is that these are just junk, remnants, but i think that these may actually be the source of new genes--genes caught in the act of evolving.
then, in a time of stress, there is a mechanism that increases the expression of these genes. i think that this must be what the article is talking about. though, i've not read the article, so i cannot be sure.
and my opinions here are not uninformed--i do this stuff for a living.
james
james
well, that's great, then! the more and varied projects you get to work on, the greater your knowledge-based will become and the more connections you'll be able to make in the data. good luck!
james
bah! can you tell i disagree?
james
as i said, it depends on what you want to do and what you want to get out of it. anyone with a narrow view and some programming abilities can generate data. it's the interpretation of the data that gets the discoveries...
I'm a computer science student, and i am enrolled in a coop program. Having basically no knowledge in biology, I have been able, within 3 months of work, to learn everything i needed and to produce enough results/data to create 3+ M.Sc./Ph.D wet lab projects.
good for you (i really mean that!). but can you analyze the data you generated and make useful conclusions that further the understanding of that particular project/field?
what you describe falls into the type 1 category i mentioned...and don't get me wrong--it's a very important category. it's just not what excites me...i'm more about having a relevant biological question (what is the function of gene x? is gene y involved in disease z? etc.) and then figuring out what tools i need (and create them if they're not already extant) and what data sources i need to mine. i can pump out gigs of data, but, in my mind, it's all about what useful information one can glean from it, not about how much you can produce. (and i'm not saying that that is what you were saying)
the more projects you become involved in, the more your knowledge base grows...be sure to have the scientists you're working with explain the _entire_ project to you so that you can see the bigger picture of what they are trying to accomplish...and maybe you'll be able to bring a different way of thinking to the project...it's not all about 1's and 0's.
james
i guess it depends on what it is you really want to do in the field..._very_ basically, there are two areas in bioinformatics: 1. the programmer who creates (possibly enterprise-level) tools as directed by the needs of the scientists and, 2. the bioinfomatics researcher/scientist who also develops tools at need, but also analyzes the data and makes conclusions and uses those conclusions/interpretations to guide wet lab work. and then, the results from the wet lab work come back to the bioinfo scientist who then incorporates the data to refine their ideas or to develop new ones which then go back into the lab. it's a very nice positive feedback loop when it works.
i fall in the latter category, which i like to call "genome hacking." the programming focus is to get the data and process it rather than making a tool that looks pretty, is user friendly, etc.
what i have found most useful in this regard is an extensive background in molecular/cellular biology (i have ~10 years of wet lab experience interspersed with my bioinformatics work (i've been full time bioinfo since '95/'96)). since molecular/cellular biology data is inherently noisy, i find that experience actually working with it and interpreting it has a profound impact on how i do my computational research as not only do i know what the wet lab is capable of doing, but i am also able to analyze wet lab data and make informed decisions based upon it...many times, this noise i spoke of has a story to tell...and sometimes it does not. it is experience that allows one to make the differentiation.
as to the type 1 bioinfo type, i always think that it is a good idea to have a working knowledge of the type of data you're processing--not just the form the data take (ie this is a text file, this is an image, etc.), but rather "this is a DNA sequence that may have errors in it and i need to be aware of that and know the types of errors that can occur so that i can include provisions for that." of course, it's more complicated than that, but i think you get the idea. of course, the best way to learn this is by doing...reading some basic molecular biology texts wouldn't hurt either. ;)
james
ummm...the one page document contains the instructions on how to set up the macs for clustering. the 200 page document contains the instructions on how to do it for linux.
why, in god's name, would you want to follow instructions that are 200 times longer?! call me crazy, but i'd rather spend the extra time actually _using_ the clustered machines rather than simply setting them up.
james
because it's XML.
james
"The 'reporter' tag defines a sequence of codons (the four amino acids that comprise DNA)"
sheeesh! can't they even get the basics right? a codon is a unit of three nucleotides that encode a single amino acid (there are three out of the 64 that do not code for an animo acid, rather, they code for the translation stop signals).
four nucleotides comprise DNA. there are 20 amino acids.
this type of error is shameful.
james
Describes a perl script that can create a virtual DNA and corresponding protein sequence with user-defined characteristics such as choice of localization (secreted, nuclear, transmembrane, etc.) and presence of particular protein domains. Resulting sequence would be considered real and coding by most prediction methods. This will allow you to file composition of matter patents on your very own custom genes.
any unsolicited emails can be olfactorily filtered by the scent of a freshly opened tin of Spam. james
Maybe I'm just being overly paranoid, but this really seems like something which shouldn't be messed with. I've heard people say that Cancer and AIDS/HIV are nature's way of dealing with overpopulation.
Sure, but if you or a loved one had either disease would you refuse treatment? The same thing could be said about polio, small pox, and any other disease. The human race got to where it's at now by overcoming these obstacles...we grow by defeating that which wants to kill us.
You take antibiotics for granted. In 10-20 years your kids will take gene therapy for granted...
Besides, I think that the human race itself will probably be its own population control--wars, famine, etc. I hope not, but...
james
The suggestion is that Celera have been doing only this, filing patents solely based on sequence homology.
Yes, this is true though they're using more than simple sequence homology when they have it. Plus, they have a big effort to obtain the actual physical entity: a full length cDNA clone (this can be a very difficult and very costly undertaking). But I think it's OK since they do have to back it up with "proof" within a year or so. I'm sure they'll let the vast majority of the applications lapse. Of course, I could be wrong.
It should also be noted that other companies can work on the same clone up until the time that the actual patent issues. Or so I've been lead to believe.
james
To discover the make-up of a protein, with a sequence of a previously unknown type, which blocks incoming self-destruct signals inside the cell, and then to prove beyond doubt that this is what the sequence does, as jrg's team did -- that seems to me a lot more impressive than just going through data automatically, and identifying everything that looks like something somebody else has seen before.
Funny you should mention the automatic data analysis because I do some of that now, having moved entirely out of the lab and fully into computational biology/bioinformatics. In this case, I feel there are merits to obtaining provisional applications based only on computational evidence. You don't simply have to rely on strict sequence homology, but you can use HMMs, motifs, and any number of other computatioanl methods to try to identify a putative function for a novel gene. Note that this is for the provisional application only and it must be backed up within a year with actual "proof" for the application to proceed. Also note that the above mentioned patent is a continuation of previous ones that we filed when we did not yet have a full length clone, but we did know a good amount about its biology...
There is simply way too much data for laboratories to keep up with--the only feasible method of sorting it out is to analyze it computationally to try to categorize novel into genes into existing families and thus have a narrower focus when it comes time to prove your prediction with actual cellular biology experiments. Helps to save time, effort, and money.
james
I checked out the first link and from what little I read, they have identified a novel sequence which is homologous to a known gene family that has some known functions (the prior art). Therefore, the utility of the discovery is clear. In addition to this they have provided tissue distribution and show that many of the tissues the gene is expressed in are associated with the particular disease states that the gene family is.
O 1&Sect2=HITOFF&d=PALL&p=1&u=/netahtml/sr chnum.htm&r=1&f=G&l=50&s1='5712381'.WKU.&OS=PN/571 2381&RS=PN/5712381
Why not check out one of mine:
http://164.195.100.11/netacgi/nph-Parser?Sect1=PT
james
I'm a scientist, not a patent laywer, so gross oversimplifications follow:
When one files a provisional patent application on a gene, one is not applying for a patent on the gene itself, but rather on a _use_ for that gene. For example, "gene_x for use in treating osteoporosis." You then usually have a year to prove your claims by providing data or at least showing that you are actively working on it rather than just sitting on it. If another company or academic institution finds that gene_x is actually involved in cancer, then they can pursue gene_x for use in treating cancer and you're SOL (shit outta luck).
The key is proving utility and this is why ESTs are unpatentable en masse. The patent office has set a very high threshold for proving utility when it comes to biotech patents; I've personally had to field questions from patent examiners and they can be brutal--for good reason: now that cDNA and sequencing technologies are so ubiquitous, pretty much anyone can generate sequence data, but not everyone can prove utility; proving utility requires time and money and, in most cases, innovation. A provisional application is as they say, "a stake in the ground" showing that you were here first--you identified this gene and you think it is involved in process X. If you can prove this, the patent may be issued to you.
I personally feel that the genome should be open source (and it will be), but I do feel that people who mine the data in clever ways and find interesting things should have legal protection for their discoveries. And let's face it: many drugs on the market are there due to the research of for-profit companies (they have bigger resources, harder timelines, etc. than academia) usually in collaboration with academia. If companies cannot profit on drugs they discover, then I wonder how much the rate of discovery of new drugs for treating diseases will decrease...
As for the whole Celera deal...I don't understand what people are getting so huffy about. I've spoken to Ventner and Celera and they've always planned to release the entire genome to the public, gratis. True, they aren't going to release it as soon as they get it (nor will they release all the work and analysis they've done on it), but the sequence itself will be released. Nothing they do with it will prevent academics from doing research...it may simply limit what other companies can do. It may be a moot point though since Celera at one point did get some DOE funding and this might make the United States Government a co-applicant of these patents. Co-applicants can make any patented claims public domain and the USG is likely to do this depending on how it all works out.
That's my 2 cents on a very complicated issue,
james