I recently gave a lecture on research methods to Siobhan Rockcastle’s ARCH 410/510 Human-Centric Environments course at the University of Oregon. Choosing the design of the UO’s Lillis Hall (ca. 2000) and BioBE’s subsequent sampling campaign (2010) as a case study, I had an opportunity to tell stories about both the architectural and energy design process and the later microbial research of one of my favorite buildings.
Lillis has been a rich vehicle of learning. I was reminded of the variety of design studies ESBL performed – site analysis, physical and computer simulations, natural ventilation controls sequence specification, and full-scale prototypes (the Energy Studies in Buildings Lab is the architectural side of BioBE) – and the diverse data BioBE collected in our first large scale project – bacteria communities using 4 methods of collection, architectural design and construction documents, building controls systems trends, Registrar class data, field observation and audits, sensor data acquisition, photos, and more became part of the rich dataset.
I took away from the lecture a renewed appreciation for case studies such as Lillis. As a design project it incubated several long-term and fruitful partnerships within the building industry. With regards to BioBE, and besides contributing 3 papers to the relatively new (at that time) scientific literature in the field, Lillis was a tremendous way during the Center’s launch to get architectural and microbial ecology experts to understand each other. It provided a wealth of learning opportunities at many levels for the scientists and students alike. In my mind, these are critical-path collaborations towards solving our grand challenges.
BioBE researchers spend quite a bit of time writing grant proposals. In particular, over the last few months I (Sue Ishaq) have been co-writing proposals which expand our understanding of indoor lighting on human health and behavior, the indoor microbiome, and energy usage in buildings. These project proposals are collaborative efforts between BioBE, Energy Studies and Buildings Laboratory, and the Baker Lighting Lab. We’ll have more updates in the next few months as those are reviewed.
Siobhan “Shevy” Rockcastle, Chair of the Baker Lighting Lab, and I have been brainstorming ideas, and today I went over to the Baker Lab to check it out in person. The Lab is decorated with concept-design lighting projects from previous students, which are not only beautiful, but extremely creative. Here are a few of my favorites!
Looking to participate in research? Amir Nezamdoost, a graduate research fellow in Energy Studies in Buildings Laboratory, is collecting survey data for his PhD dissertation. Amir’s dissertation is about developing a rating system to score the quality of window view based on human visual perception and judgmental preference. He is running a piloting survey online, and looking for participants to try it out! The survey should take less than 5 minutes.
AMIR NEZAMDOOST, LEED Green Associate
Ph.D. Student in Architecture | University of Oregon
Graduate Research Fellow | Energy Studies in Buildings Laboratory
Committee Member of LEED IEQ Technical Group
Advisory Committee Member of Illuminating Engineering Society
Student Member of the IALD, ASHRAE, SBSE, SLL firstname.lastname@example.org
Several months ago, I connected with Dr. Kevin van Den Wymelenberg at the University of Oregon about the interaction between indoor microbiology and indoor chemistry, and several other common interests. We also discussed the possibility of having me visit the University of Oregon during my fall semester leave from the University of Texas at Austin. I visited the University of Oregon for 12 days in October, and I am so glad that I did.
Kevin walks in fairly large circles while wearing a lot of hats at the University of Oregon. He is the Director of the Energy Studies in Buildings Laboratory (ESBL), and Co-director of the Biology of the Built Environment (BioBE) center, and this summer and fall has been serving as the interim Head of the Department of Architecture. The BioBE center is a unique initiative funded by the Alfred P. Sloan Foundation’s Microbiology of the Built Environment (MoBE) program. Prior to my visit I was already aware of, and impressed by, how BioBE has effectively built a community of scholars that bridge two disparate disciplines; architecture and microbiology, in a transdisciplinary manner. That impression was underscored and reinforced during my visit.
I delivered two seminars during my visit, both in the School of Architecture and both entitled Living in a Material World – How the Surfaces that Surround You Affect What You Breathe. One of these was delivered on the main campus of the University of Oregon in Eugene, and the other to the School of Architecture’s program in Portland. The latter also contains the ESBL’s climate chamber facility, which is a highly-controlled and impressive instrument for studying indoor environmental quality. The seminars were well attended, and the questions and discussion which followed each were both insightful and creative.
I also interacted with numerous individuals during my visit, most of whom are dedicated to improvements in building energy consumption, as well as expanding knowledge related to indoor air quality and particularly microbiomes of buildings. We discussed their ongoing projects, including one involving the effects of lighting on bacteria in residential dust and funded by the Alfred P. Sloan Foundation (Link: ) and another on home weatherization funded by the U.S. EPA. This led to a broader discussion about future research with BioBE’s Sue Ishaq and ESBL’s Jeff Kline, about the impacts that home weatherization and operation, as well as human behavior can have on indoor chemistry, for example, the use of air fresheners or cooking, which release highly reactive unsaturated organic gases and particles to indoor environments. We even wondered whether indoor chemistry would impact microbial communities, and our discussions sparked a few pilot projects between myself, BioBE/ESBL, and one of my colleagues at the University of Texas at Austin.
There were many other great interactions during my visit to the University of Oregon, including one with Isabel Rivera, a Ph.D. student who is doing important work on indoor air quality and thermal comfort in elementary school classrooms in Chile. I conveyed to Isabel some of the findings from my own studies on indoor air quality in schools in Texas, as well as ideas for low-cost measurements of pollutants emitted from unvented space heaters in Chilean schools.
Not all of the time during my visit was for work. Eugene is a great city for walking. The weather was fantastic and fall colors were in the height of their flare. I walked everywhere. I walked and walked and walked, then rested a bit and walked some more. A highlight was visiting the historic Mims home near downtown Eugene.
Another highlight was attending an Oregon Ducks football game (they beat the Utes). My ex-PhD student, Elliott Gall, now an assistant professor at Portland State University, came down to Eugene for the game. Great fun – and I have learned to appreciate both the quack attack and the hand gesture “O” (for University of Oregon). All in all, I was impressed by everything Eugene, at the center of which was the University of Oregon and the people at ESBL and BioBE.
We attended the Oregon Climate Change Research Institute Mini-Conference at Oregon State University on 10/18/17. This half-day event featured 4 minute lightning talks on a broad gamut of climate change related topics, including presentations on atmospheric modeling, meteorology, ocean and tidal science, vegetation impacts, public health, and legal efforts, and featuring an extended “conversation” on ethical and moral issues. Jeff Kline of BioBE presented a short talk titled “Consuming and Producing Climate change Research” which covered the building science and microbial ecology aspects of our work.
We do a lot of Illumina-based metabarcode sequencing here at the BioBE center. Sequencing is getting cheaper, and the amount of data you can get from a sequencing run continues to increase, but not at the same rate: it is now becoming more and more common to sequence samples across multiple sequencing runs, because a single run does not provide the necessary sequencing depth.
The field, as a whole, is still trying to work out how combine samples from different sequencing runs: because the error rates and read distributions tend to be specific to a given sequencing run, it can be difficult to distinguish between run effects and biological effects.
We’ve recently run across an interesting case, while working to improve our bioinformatics pipelines.
It is common for sequencing facilities to spike in Phi-X DNA to add heterogeneity to the library being sequenced; this heterogeneity prevents synchronous fluorescence from any given base overwhelming the sensor (Phi-X reads are removed bioinformatically, generally by the sequencing facility). There is, however, a more sequencing-efficient way to introduce heterogeneity into your library: variable length spaces between the Illumina adapter and the target sequence. This method doesn’t “waste” sequencing on Phi-X, but still handily prevents synchronous fluorescence. The problem is, sometimes those spaces may not be fully removed before data processing.
In collaborating with colleagues to test various options for merging data from distinct sequencing runs, we were working with some problematic data that included samples re-sequenced in two different Illumina MiSeq runs. We discovered that they had such heterogeneity spaces that had not been removed by the sequencing facility. This didn’t matter at all when processing with QIIME and uclust, because the 97% OTU radius was enough to “lump” all of the spacer sequences into the same OTU, but when working with denoising tools that infer exact sequence variants (ESVs), like DADA2, it altered the ability to recognize that the dominant sequences from the same sample in the two different runs were the same.
There are several potential solutions to this problem, but the best one is to always make sure you understand your data fully, and remove any potential sources of artificial variation before inferring sequence variants or picking OTUs. Usually, that means searching for and removing the PCR primers from each sequence, along with any sequence behind them — there are many programs out there with this functionality, including the FastX toolkit, trimmomatic, and cutadapt. If you’re sequencing a variable length region, like the internal transcribed spacer (ITS) of the ribosomal DNA, this also has the benefit of removing artificial variation introduced by sequencing past the primer on the other side of the short amplicons.
Another potential solution if you’re using DADA2 is to use “100%” OTU clustering (that is collapsing all sequences that differ only be length into the same inferred variant). There is, conveniently, an option baked into dada() for that: collapseNoMismatch = TRUE. The DADA2 pipeline also did a much better job of recognizing that different sequences with artificial variation were actually the same when using pool = TRUE, although pooling all samples for sequence inference is likely too computationally intensive to be a viable solution.
Additionally, Paul McMurdie points out that we can look for irregularities early on with DADA2:
Another way to note this early in your process is to check that the error rates look reasonable for your platform/amplicon, e.g. if you had previous successful runs for that amplicon and seq platform, you could check that the error profiles are not wildly different. If they are, you usually have a problem with trimming.
I am, however, not able to see a clear signal of the heterogeneity spacers in the error profiles for this data. It may vary with the length and variation within the spacers — I’ll surely be adding an error profile check to my standard workflow, though.
After removing the primers and spaces, we get much better agreement between sequencing runs (although we still get 45–65% of ESVs in only one run or the other). We’re still investigating this particular issue: you can follow the ongoing discussion (and contribute!) on the DADA2 GitHub page.
Last week, the staff of the Biology and the Built Environment Center presented cutting-edge science from the Center and beyond to a group of interested practitioners. The Design Champs webinar series is intended to communicate new scientific advances in the field of indoor microbial ecology research to architects, engineers, and other interested parties. For this second seminar in the series, we had representatives in attendance from:
The group was lively, and participated in a active discussion of some of the science we’ve been doing at the BioBE Center lately. In particular, we briefed them on some thoughts on hygiene that we’ve been having lately, and then discussed how that might impact the way we think about design; next, we discussed the human microbial cloud, tying the idea into the discussion of hygiene and design; this led smoothly to a discussion of some of our most recent work, focusing on the transmission of microbes to the human skin microbiome. After discussing how hygiene serves as a conceptual frame for understanding both of those studies, we went on to talk about antimicrobial compounds in built environments, and how that relates to the spread of antibiotic resistance genes.
The webinar finished with a preview of related new work — a much larger study on antibiotic resistance genes in indoor microbiota, conducted across dozen of gyms in the Pacific Northwest, and including the synergistic use of next-generation sequencing for metabarcoding and metagenomics, and targeted LC-MS/MS and intensive antibiotic-resistance culture assays in association with colleagues at Northwestern in Chicago.
The 2017 ESA meeting in Portland, OR, which took place August 6-11, created a flurry of imagination here in Eugene: Roo Vandegrift left with a large hash of approximately 275 live-tweets, Sue Ishaq left with a jumbled pile of hastily scribbled notes in the program book margins, a few of which she has expanded upon, and Ashkaan Fahimipour went away with the inscrutable expression of a mathematical modeler visualizing complex networks in their head. All three presented some of their recent or ongoing work, along with a number of other BioBE members and friends from the UO Institute of Evolution and Ecology.
The meeting started out with a number of engaging science activities, including the Field-to-Collection BioBlitz, which brought conference participants to Forest Park in Portland to collect biological samples for identification and curation. Forest Park is the largest urban forest in the United States, and the biological specimens collected will shed light on the number and types of diversity found there, as well as indicate the success of urban forests at harboring a sustainable level of biodiversity.
Ashkaan gave a presentation on Thursday on The dynamics of food web assembly: Structure, stability, and trophic cascades. The study explored how empty ecosystems acquire new species, how the food web develops over time, and how the trophic niches of those colonists can determine the total diversity of the ecosystem or weather disturbances. The large meeting room was well-attended, despite the low total abundance pictured- ecologists don’t seem to like to sit in the front rows.
In between our presentations, we filled our days by attending other talks and posters, networking events, and daydreaming about our own science. We took away valuable perspectives on newly discovered results, considerations for data analysis, or the dynamics of ecological systems, which can be incorporated into our own work to improve how we think about indoor systems and approach problem solving. You never know when a presentation on shower heads, baboons, or dormant amoeba might give you an idea which will change the way you think.
Looking ahead, we are anticipating attending a number of conferences on microbiology, air quality, building health, architecture, and ecology over the next year. Here are a few of the meetings that are already on our calendars:
Humans spend most of their time indoors, exposed to bacterial communities found in dust. Understanding what determines the structure of these communities may therefore have relevance for human health. Light exposure in particular is a critical building design consideration and is known to alter growth and mortality rates of many bacterial populations, but the effects of light on the structure of entire dust communities are unclear.
We performed a controlled microcosm experiment designed to parse the effects of filtered solar radiation on the structure of dust microbial communities.
We report that exposure to light per se has marked effects on community diversity, composition and viability, while variation in light dosage or particular wavelengths experienced are associated with nuanced changes in community structure. Our results suggest that architects and lighting professionals designing rooms with more or less access to daylight may play a role in shaping bacterial communities associated with indoor dust.