Hello
11 views
Skip to first unread message
alan2here
Jun 17, 2011, 9:04:03 AM6/17/11
to GenoCAD User Group
Hello. I'm Alan. I'm interested in this area though the amazing DNA
hacking talk http://video.google.com/videoplay?docid=-6950604815683841321
(frustratingly, recently something has happened to apparently the only
video of it) and the various TED talks about this and similar issues.
I'm mostly a computer scientist myself. Interested in both
conventional languages like C++ and areas like Cellular Automata such
as Wireworld where an object is created on a square grid where it's
shape defines it's behaviour.
I like GenoCAD's use of abstraction and I'd like to see this stuff
become more accessible.
hacking talk http://video.google.com/videoplay?docid=-6950604815683841321
(frustratingly, recently something has happened to apparently the only
video of it) and the various TED talks about this and similar issues.
I'm mostly a computer scientist myself. Interested in both
conventional languages like C++ and areas like Cellular Automata such
as Wireworld where an object is created on a square grid where it's
shape defines it's behaviour.
I like GenoCAD's use of abstraction and I'd like to see this stuff
become more accessible.
Jean Peccoud
Jun 17, 2011, 3:05:41 PM6/17/11
to GenoCAD User Group
Hi Alan,
It seems that the link to the video is broken, What is it about? Who
gave the talk and where?
What do you mean about GenoCAD being more accessible?
Jean
On Jun 17, 9:04 am, alan2here <alan2h...@gmail.com> wrote:
> Hello. I'm Alan. I'm interested in this area though the amazing DNA
> hacking talkhttp://video.google.com/videoplay?docid=-6950604815683841321
It seems that the link to the video is broken, What is it about? Who
gave the talk and where?
What do you mean about GenoCAD being more accessible?
Jean
On Jun 17, 9:04 am, alan2here <alan2h...@gmail.com> wrote:
> Hello. I'm Alan. I'm interested in this area though the amazing DNA
alan2here
Jun 18, 2011, 3:47:04 PM6/18/11
to GenoCAD User Group
> It seems that the link to the video is broken.
Yes. As I said the link dosn't currently work.
> What is it about? Who gave the talk and where?
I think it is the video embedded in this page. The page gives more
information.
http://hackaday.com/2008/01/05/24c3-hacking-dna/
The idea of abstracting DNA sequaces into named parts, containing base
pair letters (ACTG) and other components makes the concept more
accessible and easyer to engineer.
For example lets say a premoter (I don't know what the sequence for a
promoter is) is ACCCCCCGT and the whole seqance is ACCCCCCGTCCG then
the seqance can be described as [premoter]CCG. GenoCAD allows for
similar abstract descriptions.
He talks about how it's become easy to design the sequances with a
text editor or a more complex enviroment for compositing sequances in
a more abstract and usefull way (GenoCad for example) and then send
the text seqance (ACTG's) by mail order to be put into special single
celled orginisms that are designed to loose there identity when they
come into contact with DNA and take on identity described by the DNA
or alternativly to be implanted into other orginisms such as yeast
before being returned to you.
The talk also covers ideas for creating parts more like the sort of
data structures and interconnecting components used in programming.
They also show a virus engeneered to infect the bactera found in the
human gut and change the smell of the gas they produce. Making your
farts smell of bannana is quite gimiky but shows the power in this
well, they also show how several named DNA blocks go together to
create thease and talk about a conferance where teams from around the
world develop thease.
Lastly they talk for a while on ethical issues and legal frameworks.
They don't presume that the audience must be biologists or even that
this is necessarily useful for every step of the process. They talk
about other abstractions appart from GenoCad style parts such as
programming language like higher level descriptive text.
They do say that abstracting reduces efficency. For example he
describes sequences of adjacent base pairs in longer seqances (parts)
as being like functions but that sometimes two overlap preventing
changing the end of one function without also changing the start of
the next. He shows how just dividing all the functions up without
enuring parts communicate in a common way or replacing any with
standard ones or anthing else that the cells devide more slowly as
they have less efficent use of energy. The same thing applies to
computer programming but I think in both cases good abstraction is
usefull.
> What do you mean about GenoCAD being more accessible?
I remember looking at this some time ago and most of the parts had
code names. For example something like D4YC99. Now they all seem to
have names, which is great, although there still appears lots of room
for improvement. In the talk he talks about part compatibility with
each other being a bit issue.
Are there example seqances and descriptions of what they do avalible
though the GenoCad interface?
Is there any simulation features\mode allowing for testing of designs.
I suposed what I'm saying is that more accessible means that it is
easier to get started and do something usefull for a beginner.
Yes. As I said the link dosn't currently work.
> What is it about? Who gave the talk and where?
information.
http://hackaday.com/2008/01/05/24c3-hacking-dna/
The idea of abstracting DNA sequaces into named parts, containing base
pair letters (ACTG) and other components makes the concept more
accessible and easyer to engineer.
For example lets say a premoter (I don't know what the sequence for a
promoter is) is ACCCCCCGT and the whole seqance is ACCCCCCGTCCG then
the seqance can be described as [premoter]CCG. GenoCAD allows for
similar abstract descriptions.
He talks about how it's become easy to design the sequances with a
text editor or a more complex enviroment for compositing sequances in
a more abstract and usefull way (GenoCad for example) and then send
the text seqance (ACTG's) by mail order to be put into special single
celled orginisms that are designed to loose there identity when they
come into contact with DNA and take on identity described by the DNA
or alternativly to be implanted into other orginisms such as yeast
before being returned to you.
The talk also covers ideas for creating parts more like the sort of
data structures and interconnecting components used in programming.
They also show a virus engeneered to infect the bactera found in the
human gut and change the smell of the gas they produce. Making your
farts smell of bannana is quite gimiky but shows the power in this
well, they also show how several named DNA blocks go together to
create thease and talk about a conferance where teams from around the
world develop thease.
Lastly they talk for a while on ethical issues and legal frameworks.
They don't presume that the audience must be biologists or even that
this is necessarily useful for every step of the process. They talk
about other abstractions appart from GenoCad style parts such as
programming language like higher level descriptive text.
They do say that abstracting reduces efficency. For example he
describes sequences of adjacent base pairs in longer seqances (parts)
as being like functions but that sometimes two overlap preventing
changing the end of one function without also changing the start of
the next. He shows how just dividing all the functions up without
enuring parts communicate in a common way or replacing any with
standard ones or anthing else that the cells devide more slowly as
they have less efficent use of energy. The same thing applies to
computer programming but I think in both cases good abstraction is
usefull.
> What do you mean about GenoCAD being more accessible?
code names. For example something like D4YC99. Now they all seem to
have names, which is great, although there still appears lots of room
for improvement. In the talk he talks about part compatibility with
each other being a bit issue.
Are there example seqances and descriptions of what they do avalible
though the GenoCad interface?
Is there any simulation features\mode allowing for testing of designs.
I suposed what I'm saying is that more accessible means that it is
easier to get started and do something usefull for a beginner.
Liao Chen
Jun 18, 2011, 7:56:08 PM6/18/11
to alan2here, GenoCAD User Group
Hi, Alan,
Actually, GenoCAD has its own language to translate the sequences and
simulate the phenotype by connecting parts with models of biological functions using attribute grammar.
More details are available at http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000529.
Hope this helps.
Best, Chen
--
You received this message because you are subscribed to the Google Groups "GenoCAD User Group" group.
To post to this group, send email to genocad-u...@googlegroups.com.
To unsubscribe from this group, send email to genocad-user-gr...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/genocad-user-group?hl=en.
--
---------------------------------------------------------------------
Chen Liao, PHD student
Key Laboratory of Quantum Information
University of Science and Technology of China
Hefei, Anhui, 230026, China
Key Laboratory of Quantum Information
University of Science and Technology of China
Hefei, Anhui, 230026, China
----------------------------------------------------------------------
alan2here
Jun 19, 2011, 4:55:06 PM6/19/11
to GenoCAD User Group
Thanks. This looks fantastic although I wasn't able to follow all of
it.
base pairs -> parts -> genes
DNA programming has contextual elements (first few paragraphs).
GenoCad's attribute grammar is good for computer scientists and also
converts into mathematics easily.
A lexer takes in DNA (base pairs list) that is not abstracted into
parts and sugests possible part combinations. :¬o
A parser checks output from a lexer for bugs, no pun intended.
How do I get started? Can I write some attribute grammar\use GenoCad
to do or calculate something? Create a part that could be used in
somone elses part or in a cell?
it.
base pairs -> parts -> genes
DNA programming has contextual elements (first few paragraphs).
GenoCad's attribute grammar is good for computer scientists and also
converts into mathematics easily.
A lexer takes in DNA (base pairs list) that is not abstracted into
parts and sugests possible part combinations. :¬o
A parser checks output from a lexer for bugs, no pun intended.
How do I get started? Can I write some attribute grammar\use GenoCad
to do or calculate something? Create a part that could be used in
somone elses part or in a cell?
Jean Peccoud
Jun 20, 2011, 5:44:52 PM6/20/11
to genocad-u...@googlegroups.com
First of all, thanks to both Alan and Chen for engaging this conversation. This is exactly the type of dynamics we are hoping to foster.
Alan, what's your background? I assume your background in more in CS than in life sciences.
We don't have yet an easy way to have people develop their own attribute grammars in GenoCAD. We're about to release the implementation of the translation into SBML but initially the grammar will have to be developed in the backend database directly.
A way to get your hands dirty might be to start by customizing the Prolog file we have in the online supplement of the PLoS CompBio publication. In a single script you would have access to all the major concepts of the attribute grammar without having to worry about understanding the complexity of the software implementation.
Something you could also try is to install GenoCAD on your computer. This would give you a better understanding of what's under the hood. There is a preliminary version of the compiler already released. It only does the syntactic verification but not the translation yet.
Having it on your own computer would make it possible for you to develop your own grammar. We'd be happy to help in any way we can.
Do you have specific application in mind?
Jean--
Alan, what's your background? I assume your background in more in CS than in life sciences.
We don't have yet an easy way to have people develop their own attribute grammars in GenoCAD. We're about to release the implementation of the translation into SBML but initially the grammar will have to be developed in the backend database directly.
A way to get your hands dirty might be to start by customizing the Prolog file we have in the online supplement of the PLoS CompBio publication. In a single script you would have access to all the major concepts of the attribute grammar without having to worry about understanding the complexity of the software implementation.
Something you could also try is to install GenoCAD on your computer. This would give you a better understanding of what's under the hood. There is a preliminary version of the compiler already released. It only does the syntactic verification but not the translation yet.
Having it on your own computer would make it possible for you to develop your own grammar. We'd be happy to help in any way we can.
Do you have specific application in mind?
Jean--
Norwegian Igem Team
Jun 21, 2011, 12:25:14 PM6/21/11
to genocad-u...@googlegroups.com
The folowing may also be of interest to the group ( it is alos Prolog
based)
based)
http://www.aber.ac.uk/en/cs/research/cb/projects/robotscientist/results/
> --
> You received this message because you are subscribed to the Google
> Groups "GenoCAD User Group" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/genocad-user-group/-/Rk-4PSVXkREJ.
The Norwegian iGemTeam
Jun 21, 2011, 12:17:39 PM6/21/11
to genocad-u...@googlegroups.com
Link to robotics reasoning Prolog : http://www.aber.ac.uk/en/cs/research/cb/projects/robotscientist/results/
Alan Tennant
Jun 24, 2011, 9:34:00 AM6/24/11
to genocad-u...@googlegroups.com
> Alan, what's your background? I assume your background in more in CS than in
> life sciences.
Yes.> life sciences.
> A way to get your hands dirty might be to start by customizing the Prolog
> file we have in the online supplement of the PLoS CompBio publication. In a
> single script you would have access to all the major concepts of the
> attribute grammar without having to worry about understanding the complexity
> of the software implementation.
using text to describe an RBS (rule based system).
http://www.csse.monash.edu.au/~lloyd/tildeLogic/Prolog.toy/Examples/solar
And it's also a programming language and it's functional. I havn't had much
experance of functional programming (dosn't just mean uses functions)
but I understand why features like lazy evaluation would be extremly usefull.
Where is the Prolog file?
> Something you could also try is to install GenoCAD on your computer. This
> would give you a better understanding of what's under the hood. There is a
> preliminary version of the compiler already released. It only does the
> syntactic verification but not the translation yet.
> Having it on your own computer would make it possible for you to develop
> your own grammar. We'd be happy to help in any way we can.
used to represent a higher level representation for the DNA sequance than
your (blocks?) shown here.
http://img684.imageshack.us/img684/4681/unledgys.png
caption: what does this part do, who knows, onwards.
Do I understand this correctly?
> Do you have specific application in mind?
An "IF" could be usefull if it dosn't already exist, a part which
takes\watches for a signal\state change given\caused by\as a result of
another part and signals\changes a state\changes it's state accordingly
depending on the input and how that instance of the IF part is set up.
Laura Adam
Jun 28, 2011, 4:26:48 PM6/28/11
to GenoCAD User Group
HI Alan,
I am working on the compilers for GenoCAD. There is no prolog file
with the GenoCAD source code because we generate parsers only when
needed, that is, when someone wants to validate a one of his designs
from 'my design' page. For now, they are based on context-free
grammars.
We are working on using attribute grammars in GenoCAD, and similarly,
prolog compilers will be generated.
You can find a prolog file implementing an attribute grammar from the
paper http://www.ploscompbiol.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pcbi.1000529.s003
Laura
I am working on the compilers for GenoCAD. There is no prolog file
with the GenoCAD source code because we generate parsers only when
needed, that is, when someone wants to validate a one of his designs
from 'my design' page. For now, they are based on context-free
grammars.
We are working on using attribute grammars in GenoCAD, and similarly,
prolog compilers will be generated.
You can find a prolog file implementing an attribute grammar from the
paper http://www.ploscompbiol.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pcbi.1000529.s003
Laura
Alan Tennant
Jul 21, 2011, 8:50:42 AM7/21/11
to genocad-u...@googlegroups.com
ok, so http://www.youtube.com/watch?v=xJFqqxxtbRY (33:00 to 33:35) and 33:55
that's what this is for right http://www.genocad.org/design so how do I get started?
I feel the conversation was going off on a tangent somewhat. What we were talking about didn't seem to match things on the user interface in genocad.com
that's what this is for right http://www.genocad.org/design so how do I get started?
I feel the conversation was going off on a tangent somewhat. What we were talking about didn't seem to match things on the user interface in genocad.com
Reply all
Reply to author
Forward
0 new messages