Introduction to Level 6 by Tony Eng
Video Transcript: At the start of every semester, I have to send an email out to all the students that pre-registered for my class. And the problem with the interface that I use is that there’s a 900-word limit. So I typically type out a draft, and then I have to look for ways to reduce it to 900 words. I’ve got to find ways to shorten things and abbreviate things, and essentially capture the essence of what I want to tell them. The assignment for this week is something where you have to take the research that you’re working on and write a short abstract that is short, sweet, and self-contained. And the challenge for that is that you’ve got so much stuff you might want to talk about– and how do you boil it down to something that’s short? Read the assignment for further details.
The assignment for this week is to write an abstract about your research for a scientifically literate audience. In some ways, an abstract is a pitch in written form. Depending on where the abstract is used, its job is to invite the audience to find out more (by reading your paper, visiting your poster, attending your talk, accepting your paper, considering your proposal, etc). As such, not only should your abstract be short, it should be engaging and compelling. Your abstract can include elements that address the Why, the Where, the What, the How and the So What, and turns out you may have already generated much of the material that can potentially go into an abstract, depending on your discipline and what stage you are in your research. Start with the answers to the previous assignment as the starting point. You may have to edit or rewrite some of them to make the relevant, consistent and/or compatible. Add additional material until you feel the story is complete and the writing is coherent. Then condense the entire abstract down to no more than a page double-spaced.
Complete The Following Steps
And here are the component pieces (this is just for illustrative purposes, you do not have to do this):
With the explosion of research activity in genomic and proteomic bioinformatics, there is an increased demand for rapid protein sequencing algorithms, and mass spectrometry(MS) has been explored as a possible tool for aiding in this process [YaMcEn96].
Most sequencing from tandem mass spectra relies on either some form of comparison to a database of known peptides, or manual sequence inference by human analysis of spectra. Such approaches encounter difficulties when presented with the spectra of unknown and novel proteins not catalogued in a database, or with complex spectra that do not easily lend themselves to manual interpretation. A few de novo approaches exist but their performance is sensitive to noise and gaps in the dataset and their scoring methods lack a formal framework for reasoning about the answer produced.
We propose a new approach that involves a probabilistic model for peptide fragmentation, a scoring function based on this model, and a simulated annealing search based on this scoring function.
Our algorithm takes as input the original mass and the MALDI-TOF PSD mass spectrum of the peptide to be sequenced, and finds the amino acid sequence consistent with the best interpretation of the spectrum under the proposed model. If the model is good and the dataset is sufficient, then the real sequence scores optimally, and simulated annealing, under the appropriate searching conditions, will converge onto this sequence.
What we found that a simple model was sufficient to correctly predict the sequence of short peptides, and that our approach exhibited some resilience to noise and gaps in the data.
Peer Feedback Instructions
Find 2-3 friends and have them read your abstract. Ask them the following questions:
- Was the author’s abstract sufficiently interesting?
- (Did it invite the reader to want to find out more about your research?)
- Did the abstract include elements that address the Why, the Where, the What, the How and the So What?
- Does the reader have any additional constructive and specific feedback for improvement?
Finished Model 1 – Nathaniel Jones
- Building performance simulations and models of human visual comfort allow us to predict situations in which daylight-caused glare will occur based on digital building models and climate data. Unfortunately, the simulation tools currently available to architects cannot produce results fast enough for interactive use during design ideation. Recently, our lab has developed the ability to run these simulations in real-time, allowing us to provide architects with immediate feedback on the potential for visual discomfort in their designs. In this paper, we compare two simulation tools for visual comfort based on their effects on design process, design performance, and user satisfaction. We presented forty subjects with two design problems, asking them to reduce the likelihood of glare in office environments while maintaining acceptable daylit illuminance values on the work plane. Each subject analyzed one space with DIVA-for-Rhino 4.0, a commercial lighting analysis plugin for the Rhinoceros CAD environment, and one space with AcceleradRT, an experimental tool for physically based real-time rendering. Both tools provided subjects with false color luminance maps representing occupant views and corresponding daylight glare probabilities, but DIVA-for-Rhino required subjects to wait for feedback from each viewpoint, while AcceleradRT provided continuous feedback as subjects navigated the space. Using as much simulation feedback as they felt necessary during a timed exercise, we asked subjects to choose static shading device options that were appropriate for all times and positions. Subjects using AcceleradRT performed more analysis and reported higher confidence in their designs and higher satisfaction with the task. In some cases, proposed design solutions informed by AcceleradRT perform notably better than do those informed by DIVA-for-Rhino.
- Building performance simulations and models of human visual comfort allow us to predict situations in which daylight-caused glare will occur based on digital building models and climate data. Unfortunately, the simulation tools currently available to architects cannot produce results fast enough for interactive use during design ideation. Recently, our lab has developed the ability to run these simulations in real-time, allowing us to provide architects with immediate feedback on the potential for visual discomfort in their designs. However, we know little about how architects react to simulation feedback presented in real-time. In our study, forty subjects with backgrounds in building design and technology completed two shading design exercises to balance glare reduction and annual daylight availability in two open office arrangements using two simulation tools with differing system response times. Subjects with access to real-time simulation feedback tested more design options, reported higher confidence in design quality and increased satisfaction with the design task, and produced better-performing final designs regarding spatial daylight autonomy and daylight glare probability.
- Glare is a common occurrence in daylit office buildings. Building performance simulations and models of human visual comfort allow us to predict when and where daylight-caused glare will occur. Unfortunately, the software tools currently available to architects cannot produce results fast enough for interactive use during design ideation. Our tool can run these simulations in real-time and provide architects with immediate feedback on the potential for visual discomfort in their designs. However, we know little about how architects react to simulation feedback presented in real-time. In our study, forty subjects with backgrounds in building design and technology completed two design exercises to balance glare reduction and annual daylight availability in two open office models using two simulation tools with different response times. Subjects with access to real-time simulation feedback tested more design options, reported higher confidence in design quality and increased satisfaction with the design task, and produced designs with more access to daylight and lower probability of glare.
- Glare is a common occurrence in daylit office buildings. Simulations and our understanding of human visual comfort allow us to predict when and where glare will occur. Unfortunately, the tools available to architects today are too slow to be useful as design aids. We developed a tool that solves the speed problem, but it remains to be seen how architects will react to the availability of real-time glare analysis. In our study, forty subjects completed two design exercises to balance glare reduction and daylight availability in two open office models using tools with different speeds. Subjects with access to real-time analysis tried more design ideas, reported higher confidence and satisfaction, and produced designs with more access to daylight and less glare.
- Glare is a common occurrence in buildings because architects lack the tools to predict its occurrence. In our study, forty subjects designed daylit spaces and attempted to reduce glare likelihood using our fast tool and another slower tool. Subjects with the faster tool tried more design ideas, reported higher confidence and satisfaction, and produced designs with more access to daylight and less glare.
Finished Model 2 – Zsgimond Varga
Attractive colloidal dispersions, suspensions of fine particles which aggregate and frequently form a space spanning elastic gel are ubiquitous materials used as thermal insulators, catalytic electrodes in fuel cells, and in many consumer care products and food stuffs. The colloidal networks in these materials can exist in a mode of free settling when the network weight exceeds its compressive yield stress. An equivalent state occurs when the network is held fixed in place and used as a filter through which the suspending fluid is pumped. In either scenario, hydrodynamic instabilities leading to loss of network integrity occur. Recent observations using ghost particle velocimetry [Secchi et al., Soft Matter, 2014] have shown that the loss of integrity is associated with the formation of eroded channels (so-called streamers) through which the fluid flows rapidly. However, a detailed micro-mechanical model of these hydrodynamic instabilities has not yet been developed. Such a model would be extremely valuable for facilitating the design of colloidal gel products with longer shelf-lives and preventing delayed collapse within the desired time of use. In this work, we use Brownian dynamics simulations of sedimenting and hydrodynamically interacting colloids in dilute colloidal gels to examine the initiation and propagation of this instability.
In simulations, we measure the evolution of the network settling rate and identify a critical point in time beyond which the velocity grows rapidly. The rapid increase of the streamer volume in the network is shown to coincide with increasing settling rate. A simple phenomenological model is developed that describes dynamically the radial growth of a streamer due to erosion of the network by rapid fluid back flow. The model exhibits a finite-time blow-up – the onset of catastrophic failure in the gel – due to activated breaking of the inter-colloid bonds. The model relates blow-up time to the relevant dimensionless groups describing the network including: the ratio of buoyant forces to network strength, the particle volume fraction, and the strength of inter-particle bonds relative to the thermal forces acting on the particles. Simulations are used to test the model and show good agreement with the predicted dependence of the blow-up time on the physical properties of the network. The model dynamics are also shown to accurately replicate ghost particle velocimetry measurements of streamer growth. Finally, we summarize our findings in a stability-state diagram that provides insight into engineering strategies for avoiding settling instabilities in networks meant to have long shelf-lives.
Case Study: Lina Colucci On Being Concise When Pitching
WARNING: This video serves as a demonstration of the underlying ideas from this exercise applied to a (then) current MIT graduate student. The video is a little long. While it isn’t necessary to watch this video in order to progress through the next level, it is encouraged.