Definition of a "pathway"
Avi Ma'ayan
definition of a Pathway. It turns out that "pathway" means different
things to different people. Below is a draft of a definition that aims
to bridge the gap and provides a specific definition that should be
acceptable by most people:
<b>A pathway is a module containing a collection of processes that
form a common function for a common purpose represented as a directed
connected subgraph where nodes are cellular components and links
represent their regulatory relationships.</b>
Please vote, add, modify, object as you see fit.
Avi
Oliver Ruebenacker
How about: A cellular pathway is any process happening in the cell
whose participants are or are composed of chemical substances.
Take care
Oliver
--
Oliver Ruebenacker, Post-Doc Researcher
Center for Cell Analysis and Modelling
http://www.ccam.uchc.edu/
Robert Arp
inclusion in our paper.
Rob Arp
Robert Arp
may be.
Best,
Rob
-----Original Message-----
From: ontologies-of-c...@googlegroups.com
[mailto:ontologies-of-c...@googlegroups.com] On Behalf Of
Richard H. Scheuermann , Ph.D.
Robert Arp
Or, come to some agreement, through dialogue, on a canonical definition.
Rob Arp
Barry Smith
composed of chemical substances, isn't it?)
howabout
A cellular pathway is a connected sequence of two or more processes
happening in the cell and involving distinct participants
(i.e. while the participants involved in each successive process in
the sequence will likely overlap with the participants in precursor
and successor processes, there will also be some change of participants)
'Connected' then needs defining
The definition should not of course imply that we know all the
processes in the sequence -- thus the pathway representation can have gaps
BS
Oliver Ruebenacker
On Tue, Apr 1, 2008 at 10:26 AM, Barry Smith <phis...@buffalo.edu> wrote:
> last bit sounds redundant (since everything inside the cell is
> composed of chemical substances, isn't it?)
Well, I am a physicist. I see there are also neutrons, electrons,
protons, quarks and W and Z bosons in the cell, and these are not
composed of chemical substances. Or are we already assuming that these
things do not exist, because we are doing biology? On the other hand,
can an electron be considered a participant in a pathway?
> howabout
>
> A cellular pathway is a connected sequence of two or more processes
> happening in the cell and involving distinct participants
> (i.e. while the participants involved in each successive process in
> the sequence will likely overlap with the participants in precursor
> and successor processes, there will also be some change of participants)
I am not sure how to count processes, since every process consists
of subprocesses and every set of processes is a process in itself. I
think you would either have to be very specific what kind of processes
you are talking about or drop the requirement that there have to be at
least two of them. For sake of simplicity, I would consider one
process a pathway, too.
> 'Connected' then needs defining
> The definition should not of course imply that we know all the
> processes in the sequence -- thus the pathway representation can have gaps
Again, for simplicity, I would drop the connected requirement,
because we would either need a restrictive definition of what kind of
processes are allowed, or we would have different representations for
the same pathway, some connected and some not.
About sequence: Is Krebs Cycle a pathway?
Take care
Oliver
Barry Smith
>At 10:46 AM 4/1/2008, Oliver Ruebenacker wrote:
>>
>> Dear friends, dear Barry
>>
>>On Tue, Apr 1, 2008 at 10:26 AM, Barry Smith <phis...@buffalo.edu> wrote:
>> > last bit sounds redundant (since everything inside the cell is
>> > composed of chemical substances, isn't it?)
>>
>> Well, I am a physicist. I see there are also neutrons, electrons,
>>protons, quarks and W and Z bosons in the cell, and these are not
>>composed of chemical substances. Or are we already assuming that these
>>things do not exist, because we are doing biology? On the other hand,
>>can an electron be considered a participant in a pathway?
I leave this question to others
>> > howabout
>> >
>> > A cellular pathway is a connected sequence of two or more processes
>> > happening in the cell and involving distinct participants
>> > (i.e. while the participants involved in each successive process in
>> > the sequence will likely overlap with the participants in precursor
>> > and successor processes, there will also be some change of participants)
>>
>> I am not sure how to count processes, since every process consists
>>of subprocesses and every set of processes is a process in itself. I
>>think you would either have to be very specific what kind of processes
>>you are talking about or drop the requirement that there have to be at
>>least two of them. For sake of simplicity, I would consider one
>>process a pathway, too.
I think to be a pathway there has to be the minimal requirement that
at least two participants (or two cellular locations) are involved
and that at different times different participants or cellular
locations are involved -- pathways have to contain the analogue of at
least one edge in a graph, don't they?
>> > 'Connected' then needs defining
>> > The definition should not of course imply that we know all the
>> > processes in the sequence -- thus the pathway representation
>> can have gaps
>>
>> Again, for simplicity, I would drop the connected requirement,
>>because we would either need a restrictive definition of what kind of
>>processes are allowed, or we would have different representations for
>>the same pathway, some connected and some not.
Different representations is what we would have anyway,
We could drop the requirement, but the problem (of defining
'connectedness' would not go away). (Life is of course simpler if you
ignore this problem. But life is simpler if you just give up biology ...)
BS
Gary Bader
A set or series of interactions, often forming a network, which
biologists have found useful to group together for organizational,
historic, biophysical or other reasons.
(maybe it would be useful to find examples of pathways where biologists
have grouped interactions for the various reasons mentioned)
Note that Nancy Gough concurred with this definition at the meeting.
However, maybe it is not precise enough for our purposes?
Another definition, proposed by Chris Sander (personal communication),
can be paraphrased as "a pathway is a series of events, starting with an
input or stimulus and ending with a measured output or phenotype". This
interesting from an experimentalist view since by definition,
perturbation of any cellular component that affects the output, given an
input, is part of the pathway.
Gary
--
http://baderlab.org
Terrence Donnelly Centre for Cellular and Biomolecular Research
University of Toronto
ulrike...@eml-r.villa-bosch.de
I would like to clarify that a pathway is an artifical definition for
a set of reactions/processes. Within the cell there is no cut off of
processes when a pathway defines an end. More or less everything is or
can be connected. A pathway is only a definition of humans to classify
the reactions within a cell for a better understanding. So a pathway
definition depends on the context of interest and could be a set and
not only a sequence of chemical reactions, translocations or other
processes. Every "object" (molecule, electron, compartment etc.) of a
cell can be a participant of an pathway.
regards,
Ulrike
Barry Smith
>Hi all,
>I would like to clarify that a pathway is an artifical definition
a pathway is not a definition
>for
>a set of reactions/processes. Within the cell there is no cut off of
>processes when a pathway defines an end. More or less everything is or
>can be connected. A pathway is only a definition of humans to classify
>the reactions within a cell for a better understanding. So a pathway
>definition depends on the context of interest and could be a set and
>not only a sequence of chemical reactions, translocations or other
>processes. Every "object" (molecule, electron, compartment etc.) of a
>cell can be a participant of an pathway.
So a molecule can be a participant in a definition?
BS
Barry Smith
>How about the one we use in BioPAX?
>
>A set or series of interactions, often forming a network, which
>biologists have found useful to group together for organizational,
>historic, biophysical or other reasons.
The problem with this is that (some) pathways existed, surely, before
any biologists found them, or found them useful
>(maybe it would be useful to find examples of pathways where biologists
>have grouped interactions for the various reasons mentioned)
>
>Note that Nancy Gough concurred with this definition at the meeting.
>However, maybe it is not precise enough for our purposes?
>
>Another definition, proposed by Chris Sander (personal communication),
>can be paraphrased as "a pathway is a series of events, starting with an
>input or stimulus and ending with a measured output or phenotype".
Same problem here: two portions of reality could be
physico-chemically identical yet only one is a pathway because only
one has an output or phenotype that is measured
Drop that 'measured', however, and this looks good
BS
Gary Bader
Barry Smith wrote:
> At 11:08 AM 4/1/2008, you wrote:
>
>> How about the one we use in BioPAX?
>>
>> A set or series of interactions, often forming a network, which
>> biologists have found useful to group together for organizational,
>> historic, biophysical or other reasons.
>
> The problem with this is that (some) pathways existed, surely, before
> any biologists found them, or found them useful
As you have previously mentioned, pathways may be fiat objects. Do fiat
objects exist before someone finds them useful to define?
-Gary
Barry Smith
>Barry Smith wrote:
> > At 11:08 AM 4/1/2008, you wrote:
> >
> >> How about the one we use in BioPAX?
> >>
> >> A set or series of interactions, often forming a network, which
> >> biologists have found useful to group together for organizational,
> >> historic, biophysical or other reasons.
> >
> > The problem with this is that (some) pathways existed, surely, before
> > any biologists found them, or found them useful
>
>As you have previously mentioned, pathways may be fiat objects. Do fiat
>objects exist before someone finds them useful to define?
That is a deep philosophical question. But to cut it short: Did Utah
exist before the relevant bits of desert were demarcated on treaties
and maps. As a political entity no. As a biophysical entity (soil,
sand, ...) yes. The implication for pathways is, I hope, clear
BS
Barry Smith
>Avi:
><b>A pathway is a module containing a collection of processes that
>form a common function for a common purpose represented as a directed
>connected subgraph where nodes are cellular components and links
>represent their regulatory relationships.</b>
There is a problem (analogous to those I raised in connection with
earlier proposals) with the word 'represented'
Pathways do not exist because they are represented.
Rather they exist [in some sense] and it is in virtue of this that
they can be represented
I would therefore suggest replacing 'represented' by 'which can be represented'
"form a common function" seems to be potentially a bit too
restrictive -- suppose two complementary parts of a pathway each
realizes its own function ?
do we need both 'function' and 'purpose'?
Markus Krummenacker
>
> Hi all,
> I would like to clarify that a pathway is an artifical definition for
> a set of reactions/processes. Within the cell there is no cut off of
> processes when a pathway defines an end. More or less everything is or
> can be connected. A pathway is only a definition of humans to classify
> the reactions within a cell for a better understanding.
Hi all,
Regarding the question of how to determine the boundaries of pathways,
here is a paper that compares BioCyc and KEGG:
What other literature references are people aware of, which discuss
the topic of boundaries ?
--
--
Regards
Markus Krummenacker