已经用PS3运算了239个蛋白质单元（高级运算模式）

该用户已被禁言

楼主太nb了，我0个的内牛满面啊

该用户已被禁言

这个是干嘛的？？一直无解

nosa

楼主当前计算的project

每个WU得点257。计算时间在6.5小时

project 3464: 257 points, timeout 3 days, deadline 6 days

These WUs should take about 6.5 hours.


该项目内容为Project 3451-3458中新增加的一些参数去做运算。

Projects 3459-3464 are projects for the PlayStation3 Folding@Home client. These are akin to projects 3451-3458, but we're now testing some new latest-generation forcefields in addition to the inner dielectic parameters in the solvation model.

调查Project 3451-3458项目后发现该项目就是Project 3445-3450中的参数模型。总结：LZ当前正在给Project 3445-3450的项目进行新参数模型的运算。

Project 3445-3450项目介绍.大致的意思是通过模拟以下三个小肽的折叠过程得出一个精确的折叠结构定义。
他们分别是(Fs肽(一种水螺旋的片段)、protein G[一种蛋白结构]的C末端结构、泛素的N末端片段)。
简单的说，我们需要开拓蛋白质折叠更精确的模型，所以我们可以用我们的最佳模式，以了解疾病的分子机理。
以下内容出自google翻译：
复杂点说，次时代（如GPU和索尼Cell处理器）仿真平台可以实现在某些物理精度的速度时间上大幅提高。最重要的是，我们使用一种叫做广义Born近似计算所有溶剂分子的净影响，在解决的蛋白质。这样做的好处是巨大的：我们可以忽略所有的自由在我们的模拟溶剂度，极大地加快我们的计算（高达200毫微秒/天，小型蛋白质noew！）

Projects 3445-3450 are some (new!) projects for the PlayStation3 Folding@Home client. Like previous PlayStation simulations, the purpose of these projects is to test a wider range of forcefield models on a series of test peptides.

Here we are simulating three small peptide fragments which have been found experimentally to fold up into well-defined structures: the Fs peptide (a helix in water), the C-terminal peptide from Protein G (40% hairpin in solution), and the N-terminal fragment of Ubiquitin (80% hairpin in solution).



Why are these tests so important? The short answer is this: we need to develop even more accurate models of protein folding, so we can best use our models to understand the molecular causes of disease.

The long answer is this: Next-generation simulation platforms (like GPUs and the Sony Cell processor) can achieve huge speed-ups in calculation times, at the expense of some physical accuracy. Most crucially, we use something called a Generalized-Born approximation to calculate the net effect of all solvent molecules that bathe proteins in solution. The benefits of doing this are huge: we can ignore all the solvent degrees of freedom in our simulations, and vastly speed up our calculations (up to 200 ns/day noew for small proteins!)

The downside of this, of course, is that Generalized-Born (GB) approximations have some problems that we aim to overcome. Since GB models ignore the inherent granularity of water, this causes certain artifacts (such as overly-strong interactions between charged residues in proteins). These artifacts can affect the overall balance of molecular forces acting on protein, and yield results that may not be fully accurate.

Up until now, we haven't had fast enough computers to fully assess these problems. This is where simulation on the PS3 can really help. By contributing to this effort, you can help us make a significant impact in this field.

beadlezhang

我自己建立的组群．．．才运算了３个．哈

nosa

LZ加入的是3213小组。也就是中国力量小组。呵呵，国内最大的组了。

ffpffp

支持楼主~~现在自己的专业就是分析蛋白质。。。对ps3这个“鸡肋”功能突然有了好感。。。

该用户已被禁言

这个程序是ps3哪里的，life playstation那里吗？一直不清楚

nds

1星期不关机就运算这个啊

zoujilong

LZ拿PS3当超级计算机了啊

suchasplus

同加入3213
累计运算了90+