zaneveld / full_spectrum_bioinformatics Goto Github PK

https://esajournals.onlinelibrary.wiley.com/doi/10.1002/bes2.1939

Currently there are some extra levels to the file structure that make this exercise more confusing than it needs to be.

Use cartoons to show high and low R2, multiple predictor variables, and interaction effects.

Use equations similar to https://lindeloev.github.io/tests-as-linear/#:~:text=Most%20of%20the%20common%20statistical,most%20students%20know%20from%20highschool. to show mathematical relationships

https://stackoverflow.com/questions/32597209/python-not-working-in-the-command-line-of-git-bash

https://stats.libretexts.org/Bookshelves/Applied_Statistics/Book%3A_Learning_Statistics_with_R_-_A_tutorial_for_Psychology_Students_and_other_Beginners_(Navarro)/16%3A_Factorial_ANOVA/16.06%3A_ANOVA_As_a_Linear_Model

...this should be 'for line in input_file'

While there, add a generator based version

circAtlas2.0 (Wu et al., 2020) Wu, W., Ji, P., & Zhao, F. (2020). CircAtlas: An integrated resource of one million highly accurate circular RNAs from 1070 vertebrate transcriptomes
(hat tip to Sadin Safi & collaborators for finding this resource)

Discuss simulation of Erdos-Renyi and Barbasi-Alberts networks using existing code.

https://www.viprbrc.org/

The project is for now moving away from JupyterBooks, but many obsolete JupyterBooks files remain. Instead, a simple script will use nbconvert and link substitution to build index.html and other pages that Github Pages requires for a site.

https://warwick.ac.uk/fac/sci/moac/people/students/peter_cock/python/ftp

Add discussion of the notion of 'whiteness' as practice rather than race, and how invisible norms unrelated to the explicit function of the university implicitly exclude folks: https://www.insidehighered.com/advice/2017/11/10/how-whiteness-structuring-interactions-higher-education-essay

Hat tip to Paola Rodriguez Hidalgo for bringing this article to my attention:

https://www.diverseeducation.com/opinion/article/15091919/ten-ways-to-retain-faculty-of-color

There are some outstanding minor grammatical / spelling / formatting errors that need revision.

To speed up projects, add. a template script and test code script, similar to the old PyCogent template scripts, but updated for Python 3 and specific to this project.

Using example commands like:
blastn -remote -db nr -outfmt 6 -out lineage_details.nr_blast_results -evalue 1e-10 -query lineage_details.tsv.fasta

Neat example using GeoPandas: https://automating-gis-processes.github.io/site/notebooks/L3/spatial_index.html
QGIS python tutorial: https://www.qgistutorials.com/en/docs/3/getting_started_with_pyqgis.html

https://www.ajnisbet.com/blog/tufte-in-matplotlib

https://twitter.com/berondam/status/1219977765027340288?s=11

"Toxic Ivory Towers: The Consequences of Work Stress on Underrepresented Minority Faculty."
Ruth E. Zambrana.

"Written/Unwritten"
Edited By Tricia Matthew

"On Being Included"
Sara N Ahmed

"Strategies to Improve Equity in Faculty Hiring"
Needhi Bhalla
https://molbiolcell.org/doi/10.1091/mbc.E19-08-0476

"Critical Mentoring: A practical guide"
Torie Weiston-Serdan
https://styluspub.presswarehouse.com/browse/book/9781620365526/Critical%20Mentoring

"I got an error message in the second-to-last box:

TypeError Traceback (most recent call last)
in
----> 1 closest_to_B1 = more_related(B1,B2,A)
2 print(""The node closest to B1 is:"", closest_to_B1.Name)

in more_related(A, B, C)
9 if AB_MRCA == AC_MRCA:
10 return
---> 11 if AB_MRCA in AC_MRCA.getAncestors():
12 return C
13 if AC_MRCA in AB_MRCA.getAncestors():

TypeError: argument of type 'NoneType' is not iterable
"

In particular, it has data on methylation AND exposures (download is a little annoying for bulk download)
https://portal.gdc.cancer.gov/

Useful short reference: https://link.springer.com/referenceworkentry/10.1007%2F978-0-387-09766-4_402#:~:text=Definition,the%20target%20DNA%20as%20input.

https://www.socialexplorer.com/

h/t Dr. Jin-Kyu Jung

https://github.com/meagmohit/EEG-Datasets

There is no reading response link at the bottom of content/04_exploring_python/error_messages_in_python.ipynb. Additionally, on the reading response form, there is no entry for this reading.

Some useful resources are here:

https://sphsc.washington.edu/stuttering-and-fluency

Text can also be taken from this response: The goal of this advice is to minimize how much stress and cognitive load you have when presenting each slide. Basically the idea is to do a lot of work up front to make the slides easy to present, so that when you are actually presenting (in whichever format) the presentation almost runs on autopilot, and you can focus all your mental energy on strategies for delivery & connecting with the audience.

1. Design each slide so that it makes a single, simple point. You get to design your presentation, so you may as well make it as easy to present as possible. Everything on the slide should in some way support, reinforce or elaborate on that single point. Here's how you can do this starting with a (not very good) starting slide.

A typical kind of slide we all make when we first start making a presentation is a bullet-point list equivalent to a written paragraph.

Maybe our points are:
a. Sexual selection leads to the development of secondary sexual characteristics
b. Characteristics useful for sexual selection may trade off with predator avoidance or other sources of fitness
c. Parental investment theory is one explanation for sexual selection.

It would be tempting to put these 3 points on a slide, and then write a few sentences out on paper to memorize and say about each. However, this would be a big mistake - I would find such a slide very difficult to present compellingly. I'd be reading for a looooong time, and the audience would get bored. The language would often be very technical, and sound odd when you read it out loud. The audience would also have to jump between reading the written text and listening to what you're saying. If these differ at all, studies have shown it decreases comprehension over just written text or just speech alone.

Here is what I would do instead:

-- Break up our 3 main points into 3 different slides. Write the main point of each slide as succinctly as you can in one phrase in the title of each slide. If I forget what to say or what the slide is about I can always literally read the title and it should jog my memory. This also helps folks who zoned out for a minute, or who might have missed something you said.

-- Instead of adding lots of text to each slide, add 1-2 meaningful images, figures, or diagrams that directly relate to what you want to say. For example, if I wanted to talk about how parental investment theory explains cases where females have elaborate displays or compete strongly for mates, I might show in one image a case where males have an elaborate display and in the other a seahorse or Jacana bird (or another example I wanted to use) where males are more choosy and females compete for them. A few explanatory labels or phrases are OK, up to maybe two sentences. But definitely not more than that.

-- Instead of writing out text ahead of time, I would not write anything out. Then, once I was happy with my title and very visual slide design, I would just try explaining it conversationally, as one would to a friend. I'd try a few times until I was happy with what I was saying. THEN (and ONLY then) I would write down roughly what I said in order to memorize it. When we do Investigative Biology it is AMAZING what a big difference this technique makes. Pre-written explanations tend to be super-technical, don't always bring the audience along, and often aren't even that relevant to the main point. When folks have to explain what they've shown the explanations tend to be much more down-to-earth, correct, and compelling. Since everything you're going to say is now something that you would normally say in conversation (if maybe a little bit more science-y) it can be much easier to present.

-- When practicing / developing what you will say, be sure that if you are showing a figure you've explained what it is testing, what the axes represent, what any colors represent. You also want to relate what the results mean back to the overall question or biological hypothesis of the paper. If that explanation is long, its sometimes best on its own slide after the figure.

-- Again, all your slides should be strongly visual. The visuals should not just be pretty pictures or random cartoons, but specifically relevant to what you will say. So if you are talking about how sexual selection can lead to competition between members of the lower parental investment sex, you could show and very briefly discuss an example in some animal. This could still only be two sentences worth of text that you're saying on the slide. For example, "Sexual selection can lead to competition among members of the sex that invests less in reproduction. For example, in this image two male elk are competing during mating season."

2. The more technical what you are trying to say on a slide is, the more it helps to make a custom diagram or visual to help walk the audience through the slide. For example, if you just list methods as text it is extremely painful to try to read them and bring the audience along. Often a simple diagram showing the experiment or analysis really helps. One common trick is to make highlighting (big yellow boxes around key text/parts of figures/parts of tables) pop up by duplicating the finished slide with the highlighting box (use the shape tool in Google Slides/Powerpoint to make it), then deleting it from the first slide. This can help reader figure out what they should be paying attention to.

3. Just as you add visuals to support your point, you also remove visual elements that don't help communicate your point. For example, delete random decorative graphics that don't relate to the point. Remove images that you don't talk about / that aren't consistent with what you're saying. Remove unnecessary text (you can keep brief text labels or short statements, but if you see a paragraph it's gotta go). Stretch all images (in Google slide or PowerPoint hold down the shift key so they keep their aspect ratio/don't get distorted) so they fill as much of the slide as possible while still having room for your title, allowing viewers to see details. Align images so the edges match up vertically and horizontally, so viewers eyes can focus on the image and what you're saying about it, and not the arrangement of the images. You can use proportional stretching (stretching with shift held down) and cropping to help the images line up nicely. There's also an 'Align' button where if you select multiple images you can have the program line up their top edge.

4. Carefully check each slide to make sure that it logically connects to the slide before and after it, and you know how you are going to connect the two. Often I have to rearrange slides to make the transitions easier for me when I present. In other words, the exact same slides can be much harder to present if they are in the wrong order. Note that the order you present results in doesn't necessarily have to be the order they are presented in the paper, if another order would make for a less stressful or more dramatic presentation.

5. Look for moments of drama. Often papers try to answer big questions. Results may pose additional questions. Really building these up can make for fun moments. For example, in the elephant tusk paper we saw, the researchers had lots of control experiments to show it really was that tusks were shrinking, not that all of the elephant was shrinking. So I worked with that group on transitions to those experiments that made sense and would be fun/dramatic to present: "So we've seen that tusks are getting smaller. But could that just be because the whole elephant is smaller? In other words, is selection really acting on tusks specifically? To address this, the researchers tested the hypothesis that elephants were getting smaller overall in response to poaching. (transition into slide on comparing elephant size over time in elephants).

Chapter Title \t Notebook \t NBViewer \t Colab \t Python Skills \t Statistical Skills \t Biological Skills

http://bioinfo.usu.edu/pySeqRNA/

This text needs to come after the code for simulating a linear model with error.

Try it out: try rerunning the code several times without changing any parameters. Does the shape of the distribution change, or stay exactly the same?

Try it out: try increasing the scale parameter, how does the shape of the normal distibution change? Do you think a higher scale parameter corresponds to more observation error or less?

https://github.com/WheatonCS/Lexos/wiki/Python-Coding-Style-Guide

Undergraduate students really struggle with how to open scientific presentations, often starting with either a definition or something else that isn't effective at grabbing audience attention.

To address this, have an exercise where folks reading the book watch the first ~2 minutes of 10 different Kavli Frontiers of Science presentations to see how they start them out.

This will require lots of care and thought - but here's the general idea. First, review the standard DEI framework.

Privilege "White Privilege and Male Privilege", McIntosh 1988
Implicit Bias
Intersectionality
Science and Politics

After briefly considering privilege, the notion that we all have different blind spots and biases, the chapter will turn to the notion that rather than asking what is required, we should ask what we can be stopped from doing.

This will then turn to a discussion of specific policies one might check at the class, lab, division or department, school, or university level.

Ideas for action

Making our workplaces more inclusive, equitable, and diverse is a big undertaking. For most of us, there are many cultural messages that have to be chipped away at over time. There is a danger that if your metric for success is everything-or-nothing, and if you are from a majority group for whom attention to DEI may not be necessary for professional survival, you might become frustrated and give up. Therefore, in my opinion, it is a good idea to consider improving your and your workplace's DEI practices as an ongoing process. Notably, this is exactly the same way we think about science: we know we're not going to gain perfect understanding of our fields in our lifetimes, but we can build an understanding that exceeds what was known in the past.

What one practice could you begin this week (and sustain in the future) that would make your workplace more inclusive, accessible, or welcoming to all?

As an undergraduate student

Consider your own biases carefully when writing student evaluations. Research has shown that numerical scores on student evaluations often show bias against URM faculty, women faculty, and gender non-conforming faculty. Stories abound in differences in student feedback when faculty who are men and faculty who are women have (secretly) switched text-based online class sections. In comparing remarks on my own student evaluations (I am white and male) with a Black faculty member in my division, there were shocking discrepancies between how mildly students reacted to mistakes I made (video lectures were sometimes posted late), vs. the very strong and vitriolic responses to Black faculty member (who posted a single video lecture late). Of course, all instructors make mistakes, and constructive criticism is useful. However, one power that you have when filling out student evaluations (or in any other evaluative position) is to consider whether the tone, style, and intensity of comments meet the same standards of professionalism you would hope an instructor would use in evaluating your work. For example, it would be strange or downright creepy if an instructor commented on a students appearance in feedback. While your negative feedback might be valid, would you write it the same for an instructor from another background?

Consider becoming involved in campus activism

Student activism can shape campuses, communities, and indeed whole countries in important ways. It is worth a few minutes of your time to read the list of policies that have changed in countries all over the world as a result of student activism.

As one example, in the United States, most Ethnic studies and Women's studies departments arose in large part in response to sustained student pressure that grew out of the Civil Right's movement during the 1960's and 1970's. For instance, at UC Berkeley, a coalition of groups led by Ysidro Macias, Richard Aoki, Charlie Brown, and LaNada Means led a protracted student strike - even in the face of police intervention- to establish an Ethnic studies program.

Not all student activism is that dramatic, but actions that quietly establish norms of behavior are very important as well. If students introduce themselves along with their pronouns, or begin each presentation with a land acknowledgement, you can bet that faculty and other students will take notice, and perhaps become more aware of viewpoints that are not always in the foreground.

Although public universities in the United States cannot themselves - in their official functions - directly endorse particular politicians or parties, they can conduct education and outreach around important issues. They may also be able to share some of their networks of alumni contacts with student activists engaged in fighting for human rights issues. For instance, after making very slow progress in reforming university policies, student activists at Rice University led by Summar McGee bypassed university bureaucracies to raise more than $93,000 for racial justice organizations from university alumni, staff, and administrators in less than 24 hours.

Political activism

Political activism can also yield important changes to academic life and its broader social context. As a student of bioinformatics, you will be better prepared than many to use data to support your arguments. If paired with humility, open-mindedness and careful listening to those most affected by the issues you are weighing in on, this ability can enhance your ability to contribute to the political life of your community. I will direct some remarks about the specifics of how this might be accomplished primarily at the United States, as that is the system I am most familiar with.

Contacting elected representatives in the United States

It is worth considering all the levels of government (as well as other actors) that could make a positive impact on an issue that you think is important. For me, this might include my 6 city councilor and my mayor, the 5 members of my county council (which oversees the limited public health care offered in my area at the county level), my representative (congressperson), my two senators, and my governor. It is well worth a half-hours worth of effort to find the numbers for these representatives and put them into your phone. When an issue arises, you will then be ready get in touch with your representatives to help address it, and to help coordinate other like-minded activists on social media.

You may also consider the actions that might be taken by other actors in civil society: local businesses, clubs and organizations, religious communities, local newspapers or neighborhood organizations.

You might brainstorm what action each of these actors could take that would help improve your issue (e.g. by issuing a statement, contributing funds, passing a policy, or covering an issue), and how you might most effectively encourage or pressure them to take that action.

In general, your most effective actions will be coordinated with others. You may find that if you are acting alone or with a small group of people, your greatest leverage may be found with smaller actors with whom you already have some sort of relationship.

Contacting Congressional Representatives

If you have never called an elected representative before, you may be surprised at how influential a small group of activists can be on local officials. In the United States, the typical congressional office is not so much larger than an academic lab. If you have an issue of concern, you are allowed to call your congressional representative to discuss these concerns. You might ask the staff person the representative's position on a particular issue or bill, encourage or condemn the representative's planned votes or seek help in navigating a bureaucratic issue for which all other avenues have proven fruitless.

Generally, the volume of communication from the public that a politician receives is roughly proportional to the area they represent. Even members of the U.S. house of representatives may not receive a very large number of calls on any given day.

You can maximize your impact by focusing on contacting your own representatives, using personalized means of communication. Representatives will generally pay the most attention to contact from constituents within their district, since that is who determines if they will be re-elected or not. They will also pay more attention to communications that come in the form of face-to-face contact, and intermediate amount of attention to phone calls or personal letters, and least attention to online contacts, form letters, or copy-and-pasted e-mails. Organizing coordinated contacts with just a small group of friends or fellow students - especially if you give your organization a name and logo - can exponentially increase your impact. All that said, any contact is better than none.

Contacting Members of Local Government

If you live in a small to medium-sized community, and call a member of your city council, or offer a comment at a city council meeting, it might be one of the only ones they here. Mayors of smaller communities may be similarly responsive.

If you have personally been affected by an issue, you might testify at a city council meeting. Personal stories from specific people can play an important role in the outcome of local legislation.

As a graduate student or postdoc

In presentations

Ensure figures are visible to folks with impaired vision. People vary greatly in visual acuity. High contrast slides and figures, clear declarative titles conveying the main message of the figure, and plain text summaries beneath each figure can help convey the message for folks with impaired vision.

Ensure figures are colorblind safe. About 1 in 20 people are colorblind in some way, with men being most commonly affected. In addition, up to 1 in 16 men are affected by deuteranomaly. Therefore, if there is a reasonably large audience, you can almost be guaranteed that someone in that audience will not be able to distinguish e.g. red and green in your charts. A simple change is to use existing colorblind-safe palettes (such as the IBM or Wong et al., 2011 palettes).

If you have strong color preferences you can use a simulation tool like (this one to check whether or not your color scheme can be distinguished by folks with colorblindness, and adjust accordingly.

For instance, I personally often use this 4-color palette: Bright green ( RGB 146,249,11), Orange (RGB 253,173,1), Magenta (RGB 255,19,170), and Deep Blue (RGB 11,79,175).

In classes that you teach

Pay attention to course structure. In addition to being generally good teaching, having strong, clear, and well-considered structures in place for how your course runs is also an important way to promote an inclusive classroom. For example, if you don't have any particular plan in mind for how you will call on students, it is quite common for a small handful of students — who likely don't represent the full diversity of the classroom — to respond to every question.

Take special care when describing medical conditions. It is important to avoid using adjectives as nouns when talking about people with a condition: "Hrithik has arthritis" is more humanizing than "Hrithik is arthritic" because it doesn't imply that the sum of Hrithik's personhood is defined by their arthritis. (Example courtesy of the BioDiversify seminar on Disability). However, terms of self-description vary over time and between people, so some attention to current trends and individual preferences is important.

**Humanize historical figures **. Science is ultimately done by humans - with all the bizarre obsessions, frailties, odd world-views, and strange stories that go along with humanity. Presenting scientists in their whole personhood may give students permission to view themselves and their full personhood as part of the scientific project. It is also useful in adjusting notions that science is a steady march of progress advanced by geniuses. Instead, many discoveries were converged on by multiple people at similar times, and even in cases where lone individuals did make amazing contributions, those same individuals were often also wrong about a great many of their other ideas.

Discuss past and present systemic racism in science. Many influential figures in statistics and population genetics were committed and explicit racists, who provided intellectual cover for the eugenics movement. Discussing the scientific contributions of these figures without acknowledging these aspects of their legacy is guaranteed to make some students feel excluded. In contrast, reflecting on the social norms and values that led past scientists to endorse atrocities, and how they may have been influenced by systemic racism, may lead naturally to reflection on which commonly accepted beliefs we might hold that from a more informed perspective are in fact horrific. Counter-examples of scientists who pushed against social trends of their time may provide models.

Record lectures and enable closed captioning. This can help students whose first language is not the same as that used in class (by allowing them to rewatch lectures), and automatic closed captioning can provide some support to deaf or hearing-impaired students (though not replacing a professional captionist). When answering questions in a recorded video, be sure to repeat the audience question - otherwise it will not be captioned and some students may be left not knowing what question is being answered.

Normalize specifying pronouns. By stating your own pronouns at the start of your class, you open the door for students to do the same. This will not make much of a difference for many students, but for non-binary or trans students this can be an important step to ensuring they are addressed properly in class.

Land acknowledgements. Many universities in the United States are built on unceded land belonging to native peoples. Land acknowledge this history at the start of a new course section, meeting, or in email signatures. While not an end goal, taking a moment to learn about the specific history of the land on which your university sits, and acknowledging that history can be a first step towards building awareness and momentum for other actions to redress historical wrongs. If you work or study at a land-grant university in the United States, the article "Land Grab Universities" (and underlying data & code) provide an excellent starting place to identify whether land expropriated from native peoples was used as seed money to fund the institution.

Incorporation of Indigenous Knowledge and Biotechnology. Formal or western science is not the only science that has happened in human history. For example, a huge proportion of modern agricultural products are based on biotechnology developed in the Americas prior to sustained contact with Europe. Inclusion of discoveries from many sites around the world — and how those discoveries interconnect — can help tell a more complete story, while also pushing back against racist notions that only one group of people in history did much to advance human understanding.

As a member or leader of a lab or research group

Establish a lab DEI policy. This can then be posted to the lab website. There are outstanding examples available (see e.g. http://www.owlnet.rice.edu/~ac53/guiding-principles-of-diver.html)

Letters of Recommendation. Letters of recommendation are important documents that can have a major impact on careers, yet they are often subject to bias.
https://csw.arizona.edu/sites/default/files/avoiding_gender_bias_in_letter_of_reference_writing.pdf

Field work safety. People with marginalized identities may need to plan for safety in ways that those without such identities do not. Your laboratory can help support this by planning field safety. https://www.nature.com/articles/s41559-020-01328-5

As a developer of code, software, or websites

Check the accessibility of websites you maintain. A checklist for web developers is available from A11y.

Universal Design and the Cut Curb effect. Often, by making your projects more accessible to people with disabilities, you may also make them more useful in a variety of unexpected cases. For example, adding 'alt text' descriptions to images so that they can be understood by readers who use a screen reader (e.g. because they are blind or have low vision) may also help preserve meaning if for some reason an image fails to load for some users. This idea is known as The Cut Curb effect, named after 'cut curbs' added to sidewalks to accommodate wheelchairs. However, these cut curbs are also useful for parents pushing baby strollers, shoppers with full shopping carts, skateboarders, etc. Changes that you make to accommodate a vulnerable group (e.g. people with disabilities) can improve the experience for everyone — an idea thereby exemplifying the value of universal design.

Add a code of conduct to software projects on GitHub. The discussions in any online forum, including the comments on code, can easily become vehicles for harassment. Therefore, many groups add codes of conduct to help establish baseline rules for how to engage. For example, software professional Coraline Ehmke developed the code_of_conduct.md that appears in many developer profiles and GitHub projects (https://where.coraline.codes/). You can promote inclusion in software projects that you lead by adopting a code of conduct, and enforcing its provisions when moderating discussions.

Push back against hush or non-disparagement clauses If you are doing research in industry, you may have opportunities to push back against hush or non-disparagement clauses that may be used to cover up discrimination or harassment.

As a member of an academic unit (division, department, school, institute etc)

Ensure Diversity Equity and Inclusion are included in your unit's Mission Statement. This recommendation would have sounded ridiculous to me just a year or two ago, but one place to start improving DEI within a department is to ensure that promotion of diversity, equity and inclusion is listed as an important goal of your unit in its mission statement. Why on earth, you might ask, does this matter? No one in their right mind voluntarily reads mission statements, right? This is all true, but once DEI issues feature prominently in a mission statement, then that represents additional leverage in every subsequent dispute about specific issues. Once everyone agrees this is an important goal for the unit, then it naturally follows that specific steps should be taken to keep track of progress towards that goal. However, it is important that whatever appears in such a mission statement has genuine buy-in as the consensus of the division. Advertising a commitment to diversity equity and inclusion that does not exist is not fair to prospective hires that are from groups historically excluded from science.

Adopt a policy on forms of address. You walk into a classroom on the first day of class. Already students will begin to form opinions about your competence, emotional availability, and likability. If you fit the (white, male, straight, cis) stereotype of 'what a professor looks like' you may be more likely to be presumed to be competent, and presumed to be deserving of professional address and interactions even before you open your mouth. Of course, with enough mis-steps this aura of professionalism can dissipate. But this benefit of the doubt - which is often differentially accorded to instructors based on their perceived identity - is an unearned advantage. That is, it is a privilege.

Female instructors, instructors of color, and instructors who present as LGBTQIA+ (among others) are often treated by students less respectfully. If you have not experienced this yourself, and ask colleagues whether students ever treat them in a way that feels disrespectful, you may hear stories that will set your blood to boiling. In the most extreme cases this behavior is almost certainly intentional. However, in more subtle cases implicit bias may be playing a role. Such implicit biases can influence what kind of behavior is 'normal' or 'expected' in subtle ways. This puts instructors with less privilege in a double-bind: either they tolerate often intolerable treatment, or they demand the respect that other instructors take for granted and risk developing a reputation as overbearing. One small simple step that divisions or departments can consider taking to address this issue in undergraduate classes is to establish an expectation that in classes instructors should be addressed by their professional title: Dr. Khan, Professor Brown, etc. The reason we have titles in the first place is that there are certain traditional norms of behavior that govern interactions with individuals that bear them. There are simply some bawdy jokes you might hesitate to tell if the Chief Justice of the United States is setting to your left at the dinner table.

Include Diversity, Equity, and Inclusion in Merit Reviews. If your division conducts annual merit reviews, you may be able to modify the form to have a separate section for diversity, equity and inclusion work in addition to the standard Research, Teaching and Service sections. Although DEI work can be woven into those other activities, a separate section gives this work the importance it should have. Even faculty interested in working on DEI issues may feel pressure not to spend very much time on them if they think it will not 'count' towards their promotion, or worse be viewed negatively by their peers. Having a separate section incentivizes everyone to think about these important issues.

Establish policies communicating an expectation of professionalism on student evaluations. Many students may not realize that student evaluations are professional documents that — despite evidence of bias — are still used in hiring and retention, and that are often read carefully by instructors. Providing some information to students about how student evaluations are used, the literature on bias in student evaluations, and expectations for professionalism in evaluations, perhaps in the form of a brief presentation or video given before evaluations start, could be one way to try to ameliorate this issue.

Establish policies to De-anonymize threatening student evaluations. Not only are student evaluations biased against under-represented faculty, they are also sometimes threatening. Perhaps surprisingly, not all universities have established policies on threatening evaluations, and may place large burdens of proof, time, and effort on faculty who are threatened, while withholding accountability from the person who issued the threat.

Further Reading

Massey et al., Ecology Letters, 2021
Additional ideas at a divisional and lab level for improving DEI in evolution and ecology specifically are found in Table 1 of this article

Add code of conduct for community contributions

I want to let students contribute "Community Examples" to FSB.

The format is roughly:

Title
Photo
Filled Photo Release Form (https://www.washington.edu/brand/graphic-elements/photography/personal-release/)
Personal Narrative (discussing your background/identity/history/perspective/interests as they relate to science)
Code example and discussion
Body of Example
Each example, should be a short piece of code that highlights a way in which you used a particular idea in a project. It's not a whole project (unless the project is very short), but rather just a key piece.

So for example, if you built a Network-based simulation of viral spread in a population, you might discuss how you used a custom python class to represent a node in a disease transmission network. If you wrote a script to calculate hydrophobicity for a protein, you might discuss how you used a dictionary to map between amino acids and hydrophobicity scores. Perhaps you devised a unique Monte Carlo test, or used a Mantel Test to compare geographic versus genetic distance matrices.

The format should be similar to other Full Spectrum Bioinformatics chapters, and shouldn't typically be much more than a single page. The idea is that we will have several short examples student examples that get automatically included at the end of each section.

https://aegis4048.github.io/mutiple_linear_regression_and_visualization_in_python

Link or build from here: https://www.dataquest.io/blog/alt-jupyter-notebook-tips-tricks-shortcuts/

https://www.amazon.com/Practical-Statistics-Data-Scientists-Essential/dp/149207294X/ref=sr_1_29?dchild=1&keywords=Python+Data+Science&qid=1633546266&sr=8-29&asin=149207294X&revisionId=&format=4&depth=1

https://nativegov.org/news/beyond-land-acknowledgment-guide/

A counter-intuitive point about traceback in Needleman-Wunsch alignment is that the cell representing the optimal path to the current cell may not have the highest score. We encountered this in an example where the cell to the left was a -2, which produces a score of -3 (-2 + -1 gap penalty). The diaganol score (above and to the left) was -4. It was tempting to think we should follow the path to the left (arrows notwithstanding), but this is not the only possible optimal alignment. Because the diagonal move had a match, -4 + (+1 match bonus) = -3. So both moves are equally optimal. For illustration, it might be useful to see a case