Is the pharmaceutical industry ready for systems biology?

Okay, it is time to move on, to out of the blue. New semester has begun, so does the theory lunch.

I really like this talk; someone from the industry telling you what it is in the real world, the systems biology.

The introduction -
--

Is the pharmaceutical industry ready for systems biology?
8 February 2008

Bruce Gomes
Systems Biology Group
Pfizer Inc.

Abstract
When reading reviews of systems biology it would seem that this method should be immediately adopted as the new paradigm for drug discovery in the pharmaceutical industry. This methodology promises to pick the best approach for amelioration of disease pathology by choosing the optimal target for drug design; identify co-drugging strategies; find the most discriminating biomarkers; stratify patients for clinical trials; militate against toxic outcomes; and overall, reduce drug discovery attrition. Yet, the adoption of systems biology methods in the industry is cautious and slow. This talk will discuss what the limitations of systems biology are that slow its adoption. These limitations are the starting points for the generation of new approaches that are a fertile area for the interface between academic labs and the pharmaceutical industry.

Even without the large scale adoption of systems biology methodologies in the industry, these methods are being used to some extent. This talk will highlight one area of systems biology that is currently being used very successfully, namely the small to moderate sized models created to predict the behavior and properties of biotherapeutics.

--

講者是 Pfizer 裡的人，從他口中大概可以理解到在真實世界中，所謂 Systems Biology 目前的可行性及應用度到底在哪裡，的確，與在象牙塔裡學術圈子想的大不相同。

通常 Systems Biology 常被質疑的就是到底有沒有可能掌握「系統」裡所有的變因 (parameters)，這個問題通常問學術圈子裡的人，答案多是持保留態度的。但就藥廠的立場，只要謹慎地選擇「標靶」，這個目標是可以達到的。

Bruce 給了幾個數據，目前市面上 FDA 共核准了 1480 種不同的藥物販售，可是這些藥物的標靶基因總數約只有 300，也就是說儘管是來自不同藥廠的藥物，許多都是 targeting 相同的基因。

這現象是許多原因加總的結果，其中一項是成本，一個新藥從一開始的評估到 phase I，平均的花費大約是 $1 billion，就算好不容易通過上市，到 Phase IV 後仍有大約 70% 的機會這個藥競爭不過市場上同質性的藥物，never got description。正因如此再一開始選擇標靶的時候，基本上是極端保守風險取向的。藥廠只會選擇 small systems，在這種情況下，所有系統的 parameters 可以全部掌握，系統中的每個成員的貢獻或影響程度可以被計算及實驗出來。

在後半段，Bruce 也表示 Systems Biology 距完全發揮應用還有一段很長的距離，實務上仍存在著許多 holes，其中 text mining 是最大的問題，業界的各大生技公司花了非常多的心力在這一部分，但很明顯的 auto-paper-reading robot 仍是遙不可及的夢，進展十分有限。其它許多臨床上的問題，像不同個體間臨床反應不同的情況，也造成了系統解析上很大的困難。

整場 talk 的感覺，業界重實務的文化確實點到了許多學術界不會考慮的問題，對整個 systems biology 的發展評估感覺也比象牙塔裡搞 systems biology 的人自己說來得準切真實多了。