f**d
发帖数: 768 | 1 一直很感兴趣，
一直没做过这方面的课题研究
希望以后做些计算模型方面的工作来解释 实验现象
谢谢！ |
m******1
发帖数: 95 | 2 你关注dopamine rewarding机制吗？还是像我们一样关注motor learning? |
f**d
发帖数: 768 | 3 都关注，尤其是皮层STDP的普适的机制
欢迎详细把你那方面理解的讲一讲
【在 m******1 的大作中提到】
: 你关注dopamine rewarding机制吗？还是像我们一样关注motor learning?
|
n****r
发帖数: 246 | 4 我老板以前做这个的，不过现在不做了
【在 f**d 的大作中提到】
: 一直很感兴趣，
: 一直没做过这方面的课题研究
: 希望以后做些计算模型方面的工作来解释 实验现象
: 谢谢！
|
w****o
发帖数: 367 | 5 Sexually Transmitted Disease Plasticity |
k*****1
发帖数: 454 | 6 其实我一直觉得STDP在 in vivo 的状况下不存在的，应该是artifact。
试想下，现在所有的STDP results都必需要求没有noise，而且couple信号很长时间。
这在in vivo的情况下是不可能发生的。 |
r****t
发帖数: 10904 | |
m***f
发帖数: 1622 | 8 有一些IN VIVO的STDP方面发表的文章吧？
说是ARTIFACT，这个观点倒是很新颖
【在 k*****1 的大作中提到】
: 其实我一直觉得STDP在 in vivo 的状况下不存在的，应该是artifact。
: 试想下，现在所有的STDP results都必需要求没有noise，而且couple信号很长时间。
: 这在in vivo的情况下是不可能发生的。
|
n****r
发帖数: 246 | 9 拜托多读点文献
【在 k*****1 的大作中提到】
: 其实我一直觉得STDP在 in vivo 的状况下不存在的，应该是artifact。
: 试想下，现在所有的STDP results都必需要求没有noise，而且couple信号很长时间。
: 这在in vivo的情况下是不可能发生的。
|
k*****1
发帖数: 454 | 10 我本身就是做STDP simulation的，文献读的够多了，要不你给些文章的链接如何？
另外，我本身的意思，并不是说STDP这个东西实验中做不出来，也不是在 in vivo的实
验条件中做不出来，而是说，在生理条件下，STDP的protocol可能没有任何意义，
第一， 前后突触信号的coupling的时间太长。一般都是15分钟以上。试想一下，大脑
里某个回路怎么可能保持15分钟时长的信号，而且还要精确coupling,而不受到干扰呢？
第二，我到目前为止还没看到任何文章能够在存在噪音的情况下，比如说spontanius
spike, 也能够实现 STDP的。请问，一个回路怎么可能在15-30分钟甚至更长的时间内
会没有“noise”呢？如果存在noise spikes，STDP 的 efficacy 会受到怎样影响？
【在 n****r 的大作中提到】
: 拜托多读点文献
|
|
|
h******s
发帖数: 728 | 11 这个问题好像muming讨论过的，不记得他怎么讲的了，其实也是比较牵强。但是他和
Yang Dan有做过一些STDP类型学习的研究，只不过是一些特别设定的例子，不见得有普
遍性。所以我觉得，STDP有多significant，现在还没有什么证据，但是应该也不是完
全就是个没用的artifact。
【在 k*****1 的大作中提到】
: 其实我一直觉得STDP在 in vivo 的状况下不存在的，应该是artifact。
: 试想下，现在所有的STDP results都必需要求没有noise，而且couple信号很长时间。
: 这在in vivo的情况下是不可能发生的。
|
k*****1
发帖数: 454 | 12 刚刚找到一篇文章， 两个电生理的专家 Lisman 和 Spruston 的， 有兴趣的话可以看
一下
Questions about STDP as a General Model of Synaptic Plasticity
John Lisman and Nelson Spruston
According to spike-timing-dependent plasticity (STDP), the timing of the Na+
spike relative to the EPSP determines whether LTP or LTD will occur. Here,
we review our reservations about STDP. Most investigations of this process
have been done under conditions in which the spike is evoked by postsynaptic
current injection. Under more realistic conditions, in which the spike is
evoked by the EPSP, the results do not generally support STDP. For instance,
low-frequency stimulation of a group of synapses can cause LTD, not the LTP
predicted by the pre-before-post sequence in STDP; this is true regardless
of whether or not the EPSP is large enough to produce a Na+ spike. With
stronger or more frequent stimulation, LTP can be induced by the same pre-
before-post timing, but in this case block of Na+ spikes does not
necessarily prevent LTP induction. Thus, Na+ spikes may facilitate LTP and/
or LTD under some conditions, but they are not necessary, a finding
consistent with their small size relative to the EPSP in many parts of
pyramidal cell dendrites. The nature of the dendritic depolarizing events
that control bidirectional plasticity is of central importance to
understanding neural function. There are several candidates, including
backpropagating action potentials, but also dendritic Ca2+ spikes, the AMPA
receptor-mediated EPSP, and NMDA receptor-mediated EPSPs or spikes. These
often appear to be more important than the Na+ spike in providing the
depolarization necessary for plasticity. We thus feel that it is premature
to accept STDP-like processes as the major determinant of LTP/LTD.
【在 h******s 的大作中提到】
: 这个问题好像muming讨论过的，不记得他怎么讲的了，其实也是比较牵强。但是他和
: Yang Dan有做过一些STDP类型学习的研究，只不过是一些特别设定的例子，不见得有普
: 遍性。所以我觉得，STDP有多significant，现在还没有什么证据，但是应该也不是完
: 全就是个没用的artifact。
|
p*******r
发帖数: 4048 | 13 Personally I don't feel it's a good time to do modeling in this field right
now.
I feel theoretically, there hasn't been any fundamental breakthrough since
my original paper.
Experimentally, the situation is quite fuzzy. I personally think STDP is
just a facet of more fundamental learning rules.
There are hundreds of proteins in the synapse. There are likely several
levesl of plasticity/metaplasticity. Some of my current work is also
focusing on teasing apart some of the intricacies. check out the latest work
by Ryohei Yasuda.
Like others commented, there is still no direct convincing proof of STDP in
vivo (actually just not in mammalian cortex). In insects, tadpole, fish, etc
, there are. I think it's highly likely STDP has a role there, but far from
the whole story. I would say it probabably contributes 50% in all the
modified forms.
We are still in need of elucidating the fundamental learning rule of the
mammalian cortex.
I had high hopes (along with the whole field) that we will find the answer
soon when I published my original paper in 2001. But sadly I feel it has
been a bit dissapointing. I feel techniques to observe the behavior of
single synapses in intact mammalian cortex is sorely needed. That's why I
have devoted almost the past 7 years to refine those techniques. Quite
painful really as it's still quite hard.
Personally I think the promising areas of theoretical neuroscience right now
is:
Reinforcement learning and inhibitory circuitry.
Recurrent and feedback connections has been a tough nut for many years, but
if you can crack it, you are golden.
【在 f**d 的大作中提到】
: 一直很感兴趣，
: 一直没做过这方面的课题研究
: 希望以后做些计算模型方面的工作来解释 实验现象
: 谢谢！
|
r****t
发帖数: 10904 | 14 I am not working on this, but am generally interested. mind to give a link
to your 2001 paper? TIA
right
work
【在 p*******r 的大作中提到】
: Personally I don't feel it's a good time to do modeling in this field right
: now.
: I feel theoretically, there hasn't been any fundamental breakthrough since
: my original paper.
: Experimentally, the situation is quite fuzzy. I personally think STDP is
: just a facet of more fundamental learning rules.
: There are hundreds of proteins in the synapse. There are likely several
: levesl of plasticity/metaplasticity. Some of my current work is also
: focusing on teasing apart some of the intricacies. check out the latest work
: by Ryohei Yasuda.
|
r***y
发帖数: 25 | 15 If we put the neural computation into several levels based on their spatial
scales, something like single synapse, group of synapse with same channel
mechanism between the same pair of pre- and post- neurons, individual pair
of pre- and post- neurons, population of neuronal pairs within the same
micro-circuit and serve the same function (i.e. LGN --> layer 4 E-E
afference), and regional cortical connection, which level do you think we
can apply STDP during simulation? The first three, the first two, or
actually we need to break down them into individual protein pathway level? I
can't remember which level your paper refers to, it's been too long from
last time I read it. :p
Also, if STDP applies to the synapse or group of synapse level, is it
possible for us to quantify the noise (or rather, the spontaneous activity)
pattern by counting the mini-PSPs for a long enough period of time and see
whether the pattern is replicable? I guess you Plos paper was trying to
model it, any further progress?
right
work
【在 p*******r 的大作中提到】
: Personally I don't feel it's a good time to do modeling in this field right
: now.
: I feel theoretically, there hasn't been any fundamental breakthrough since
: my original paper.
: Experimentally, the situation is quite fuzzy. I personally think STDP is
: just a facet of more fundamental learning rules.
: There are hundreds of proteins in the synapse. There are likely several
: levesl of plasticity/metaplasticity. Some of my current work is also
: focusing on teasing apart some of the intricacies. check out the latest work
: by Ryohei Yasuda.
|
p*******r
发帖数: 4048 | 16 The latest I am working on is imaging dynamics of single synapses in vivo.
We and others so far have good evidence that synapses scale multiplicatively
in vivo.
There are several themes we are pursuing next:
1 there are probably several levels of set points in the synapse. For
example, a synapse could have strength A in the morning of day 1, go to B in
the evening of day 2, then go exactly back to strength A on day 3. There
could be some memory, perhaps encoded by other molecules rather than
glutamate receptor which determines synaptic strength at any given moment.
2 We are trying to do direct links between activity history and synaptic
strengths in vivo.
The hope was to use STDP as a phenomenological model at individual synapses.
But this phenomenon turns out to be too diverse and complicated in practice
. I now believe we probably need to go to mechanistic model like Hodgkin and
Huxley to settle this. We need to directly observe several "hidden
variables" akin to voltage clamp. Fortunately this is quickly becoming a
reality.
>Also, if STDP applies to the synapse or group of synapse level, is it
possible for us to quantify the noise (or rather, the spontaneous activity)
pattern by counting the mini-PSPs for a long enough period of time and see
whether the pattern is replicable? I guess you Plos paper was trying to
model it, any further progress?
Problem is there is no mature method to monitor mini at single synapses yet.
What people normally measure is an aggregate over the whole cell. I have
been trying to get technologies on this front for almost 10 years now. I
think this direction is starting to bear fruits.
Some people claim the spontaneous pattern of what a cell receives is
reproducible at least at some places and some situations. I think it's
somewhat in between. The brain uses a lot of rate coding, but also some
timing based pattern codes, especially in the form of oscillations. |
p*******r
发帖数: 4048 | 17 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.8.1586
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.125.97
【在 r****t 的大作中提到】
: I am not working on this, but am generally interested. mind to give a link
: to your 2001 paper? TIA
:
: right
: work
|
d*****r
发帖数: 2583 | |
