From: W. Trevor King Date: Fri, 21 Jun 2013 17:36:04 +0000 (-0400) Subject: hooke/conclusions.tex: Discuss frameworks and automation X-Git-Tag: v1.0~37 X-Git-Url: http://git.tremily.us/?a=commitdiff_plain;h=002873cb12a91e29b2fa7ddc13e79fa4aebf3424;p=thesis.git hooke/conclusions.tex: Discuss frameworks and automation --- diff --git a/src/hooke/conclusions.tex b/src/hooke/conclusions.tex index 36916c3..43d87a7 100644 --- a/src/hooke/conclusions.tex +++ b/src/hooke/conclusions.tex @@ -1,3 +1,23 @@ \subsection{Conclusions} \label{sec:hooke:conclusions} +My \Hooke\ work builds on previous open source development from other +groups to provide a solid, extensible framework for processing raw +force spectroscopy data. With this framework, it's easy to add +analysis plugins to support new approaches to data analysis. It's +also easy to add drivers to support new data file formats. + +In it's current state, \Hooke\ is used by a handful of people to +analyze single molecule velocity clamp unfolding experiments. Even +with Hooke helping out, the process is not automatic. Because of the +scarcity of clean curves, it's hard to set agressive margins for +automatically filtering clean unfolding sawteeth from curves that +contain other interactions. Manually identifying clean curves from +the rest is time consuming and subjective. By writing software to +objectively identify unfolding events, we can improve the quality of +the resulting science while at the same time increasing throughput. +Real world data is messy, so developing objective filtering procedures +is difficult. With \Hooke's standardized framework and broad support +for existing analysis procedures, it will be easy to apply new +objective filtering procedures to old data, and see how well the new +procedures match up with their partially automated predecessors.