Introduction

Over the past two decades, the rapid development of high-throughput technologies, data storage capacity, and sophisticated algorithms has produced substantial changes in research practices in the life sciences and medicine. In order for researchers and practitioners in these areas to take advantage of the changes in data availability, they need to have computational and quantitative skills—such as those encompassed by bioinformatics—beyond what was required of them in the past. For more than a decade, authoritative calls from a variety of professional organizations to update undergraduate life sciences curricula have stressed the importance of increasing quantitative and computational education to prepare life sciences students for 21st-century careers [1–14]. Examples include BIO2010: Transforming Undergraduate Education for Future Research Biologists, a report by the National Academy of Sciences [1]; Math & Bio 2010: Linking Undergraduate Disciplines, a joint project of the American Association for the Advancement of Sciences, the American Society for Mathematical Biology, and the Mathematical Association of America [11]; Scientific Foundations for Future Physicians, a publication from the Association of American Medical Colleges and the Howard Hughes Medical Institute [12]; and Vision and Change in Undergraduate Biology Education: A Call to Action [13] and Chronicling Change, Inspiring the Future [14]. These reports are consistent in their recommendations that life sciences majors should receive better training in chemistry, physics, mathematics, and computation because such knowledge and skills are necessary to address questions at all levels of biology. In addition, several publications have called for life scientists to develop more sophisticated data analysis and programming skills so that they can benefit from the bioinformatics revolution [7, 15–20]. These include a report from the RECOMB Bioinformatics Education Conference [7] and a recent essay in Nature [17]. Along these lines, it is important to note that in its 2015 employment predictions, the United States Bureau of Labor Statistics projected that between now and 2024, 75% of new science, technology, engineering, and mathematics (STEM) jobs will involve computation [21, 22].

On a national scale, a number of innovative programs have been created to address the need for increased computational and quantitative training in biology. These include the Science Education Alliance-Phage Hunters Advancing Genomics and Evolutionary Science (SEA-PHAGES) program [23], Genomics Education Partnership (GEP) [3, 24], Genome Consortium for Active Teaching (GCAT) [4], GCAT NextGen Sequencing Group (GCAT-SEEK) [25], and Genome Solver [10]. On a local scale, we are aware that the number of courses and meetings focused on undergraduate bioinformatics is increasing and that many existing courses have been modified to include bioinformatics. However, despite these efforts, the adoption of bioinformatics is often limited either to a small number of institutions or to particular courses within a given curriculum. To take an example, SEA-PHAGES engages first-year undergraduates in a genuine research experience and includes substantial bioinformatics analysis as part of its year-long curriculum. As reported in 2014, the program had been used at seventy-three institutions by 4,800 students over five years [23]. Although the reach of SEA-PHAGES is impressive, it impacts only a small fraction of the thousands of U.S. institutions that offer a biology degree and the approximately 110,000 students who now graduate annually with biological and biomedical sciences degrees (not to mention the 220,000 more who graduate from health and related professional programs) [26]. In addition, because the program is designed for first-year students, it does not necessarily impact the rest of the curriculum at a particular institution. The reaches of the other mentioned programs are similar to, or not as extensive as, that of SEA-PHAGES. Despite these large-scale efforts and those of independent faculty to build resources for bioinformatics education, bioinformatics is not yet a standard component of life sciences education [15].

Several groups have made independent efforts to develop or describe bioinformatics curricula, but they have primarily focused on training undergraduate and graduate bioinformaticians, not undergraduate life scientists. For example, the Curriculum Task Force of the Education Committee of the International Society for Computational Biology (ISCB) surveyed directors of bioinformatics core facilities in Europe, Israel, the United States, and Canada about the skills needed for success in the field of bioinformatics and what skills were lacking in the bioinformaticians they recently hired. Based on the results, the Task Force developed a set of core competencies for bioinformatics professionals [9, 27]. They also described three professional roles that require bioinformatics training and the different but overlapping competencies required for individuals in those roles. Since their publication, the ISCB competencies have been used to establish and refine bioinformatics degree and certificate programs, tracks, and specializations in the U.S., Africa [28], Australia, and the United Kingdom. As another example, Koch and Fuellen described the bioinformatics curricula at various German universities at both the graduate and undergraduate level [29]. However, a few authors have considered the specific bioinformatics needs of general life scientists. For example, Maloney et al. discussed the importance of incorporating bioinformatics into undergraduate biology curricula and presented examples of where this has been done successfully [30]. The East Asia Bioinformation Network developed a bioinformatics skill set for biology students in the developing countries that make up the Association for Southeast Asian Nations (https://eabn.apbionet.org/3eabn08/docs.shtml). Finally, Tan et al. proposed a minimum bioinformatics skill set for life sciences curricula (graduate and undergraduate) in developed nations, formulated from discussions at a conference focused on education in the Asia-Pacific region [8]. Thanks to developments such as these, the state of integration of bioinformatics into life sciences education is maturing.

The Network for Integrating Bioinformatics into Life Sciences Education (NIBLSE, pronounced “nibbles”) is using an evidence-based approach to expand and promote the integration of bioinformatics skills and knowledge into undergraduate life sciences education [31]. NIBLSE is a National Science Foundation (NSF) Undergraduate Biology Education Research Coordination Network (RCN-UBE) that was formed in 2014 to build on and expand the curricular developments mentioned above. Since further integration of bioinformatics into undergraduate life sciences education will require the participation of non-expert faculty, one of the goals of NIBLSE is to provide resources (e.g., curricular materials and assessments) and support to faculty interested in expanding the integration of bioinformatics within their departments and programs. Note that NIBLSE is focused on the integration of bioinformatics into undergraduate life sciences curricula, which is related to, but distinct from, the curricular requirements of a “pure” bioinformatics degree. The goals of NIBLSE and the ISCB Task Force mentioned above are thus distinct.

In order for NIBLSE to provide curricular resources and assessment tools that align with the needs of undergraduate life sciences students, a set of core bioinformatics competencies for the training of these students is needed as a framework. To that end, NIBLSE recently conducted a survey (hereafter the “NIBLSE survey” or just the “survey”) that targeted U.S. biologists. The survey was designed to assess the importance of bioinformatics and bioinformatics skills for undergraduate life scientists, and it included questions about teaching bioinformatics to this group of students and the challenges involved in doing so. In addition, to get a sense of which bioinformatics skills faculty are covering now, respondents were asked to submit syllabi from their courses that incorporate bioinformatics. The analyses of the survey results and syllabi were the subjects of a national NIBLSE conference in August 2016. Outcomes of the analyses and conference are described below and include a proposed set of core bioinformatics competencies for undergraduate life sciences majors. A separate paper addresses the barriers to integrating bioinformatics into life sciences education and NIBLSE’s recommendations for overcoming them [32].

The NIBLSE Bioinformatics Core Competencies for Undergraduate Life Scientists (hereafter the “NIBLSE Core Competencies” or just the “Core Competencies”) proposed here are intended to serve as a guide for institutions as they integrate bioinformatics into their own life sciences curricula. To facilitate this necessary and important change, NIBLSE is collecting and helping to develop curricular resources for faculty and will soon be developing assessment tools and faculty development resources that are aligned with the Core Competencies.

Materials and methods

Survey results and analysis

A total of 1,260 responses to the survey were received (n = 1,260), 82% of which came from unique links (Fig 1). Survey results were analyzed by a subset of the NIBLSE CCWG, the interdisciplinary Core Competencies Team (CCT). Differences in Likert responses by self-reported demographics were determined by applying a two-sample Kolmogorov-Smirnov (KS) test to pairwise comparisons of groups. The KS tests were implemented in scripts written in R; these scripts are available at the NIBLSE repository on GitHub [34].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Geographic distribution of NIBLSE survey respondents.

The location (city/state) of each response to the survey was obtained using e-mail and/or IP addresses. The distribution of responses for the contiguous U.S. is shown (n = 1,081). A light circle represents one response at a particular location; a darker circle represents multiple responses at the same location (the darker the circle, the more responses). Note that the legend applies to the states themselves—e.g., there were more than seventy-five responses from California—and that there are no states with no responses. Responses (not shown) were also received from Alaska, Hawaii, Argentina, Australia, Canada, Denmark, France, Italy, Korea, New Zealand, Norway, Puerto Rico, the Republic of Poland, Switzerland, and the United Kingdom.

https://doi.org/10.1371/journal.pone.0196878.g001

Using public sources [35–37] and information from MDR, we estimate that there are currently between 75,000 and 100,000 biological sciences faculty in the U.S. Given our overall sample size (1% to 2%) and the number of responses per question (between 970 and 1,009), we estimate that the mean margin of error for the survey questions described in this paper is ±3% at the 95% confidence interval.

Results

Demographics

The group of survey respondents was distributed fairly evenly between males and females as well as across institutions by Carnegie classification, by measurement of institution size (total students, total undergraduates, majors, and faculty), and by self-reported minority-serving status versus not minority-serving (Fig 2). Respondents were from all fifty states with the distribution of responses roughly matching the population distribution of the country (Fig 1). The respondents were predominantly white and non-Latino/Hispanic; for the vast majority, a PhD was the highest earned degree (Fig 2). The level of bioinformatics training varied widely among respondents, including no training, self-taught, short courses, undergraduate and graduate classes, and graduate degrees (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Demographics of survey respondents.

The number of responses (y-axes) for each of the demographic variables (x-axes) on the survey, as follows: (A) Gender. (B) Race (People of Color and White); four categories in Race—American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander—were combined into People of Color (POC) due to very small sample numbers for each category. (C) Ethnicity (Hispanic-Latino, non-Hispanic/Latino). (D) Minority Serving (whether or not the respondent’s home institution is classified as minority-serving). (E) Highest Degree (highest degree earned: Bachelor’s, Master’s, Professional Degree, PhD). (F) Year Earned (year that the highest degree was earned; responses were grouped in the following bins: Before 1980, 1980 to 1989, 1990 to 1999, 2000 to 2009, and After 2009). (G) Training (level of bioinformatics training: None, Self-taught, Short workshop, Undergraduate/PostBacc training, Graduate class, and Graduate degree); four categories in Training—Undergraduate course, Undergraduate certificate, Undergraduate degree, and Post-baccalaureate certificate—were grouped together into “Undergrad” (undergraduate/post-baccalaureate training) due to small sample numbers in these categories. (H) Carnegie (Carnegie classification of the respondent’s home institution: Associate’s, Baccalaureate, Master’s, Doctoral). (I) Total Students (total number of students at the respondent’s home institution). (J) Total Undergraduates (number of undergraduates at the respondent’s home institution). (K) Undergraduate Majors (number of undergraduate majors in the respondent’s home department). (L) Faculty (number of faculty in the respondent’s home department).

https://doi.org/10.1371/journal.pone.0196878.g002

Ratings of bioinformatics skills

As discussed in Materials and Methods, the survey asked respondents to provide Likert-scale responses about the importance of fifteen bioinformatics skills, S1 to S15, in undergraduate life sciences education (see that section for a description of each). Analysis of this data focused on determining, using a two-sample Kolmogorov-Smirnov (KS) test (see Materials and Methods), whether the Likert responses differed significantly across demographic variables. Results are summarized in Fig 3 and Table 1. Skills that received the highest mean responses overall included S1 (Role), S3 (Statistics), S4 (Access genomic), and S5 (Tools genomic) (Table 1). As detailed below, the perceived importance of some skills varied based on the demographics of the respondents.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Summary of bioinformatics skills ratings.

The total number of responses (y-axes) by Likert-scale rating from 1 to 5 (x-axes)—1 being “Not at all important” to 5 being “Extremely important”—for each of the fifteen survey skills, S1 to S15, labeled in sequence from (A) to (O). As discussed in Results, these skills were divided into two broad categories: skills that just required familiarity (“knowing” skills: S1 to S4, S6, S8, S10), and those that required direct engagement (“practicing” skills: S5, S7, S9, S11 to S15).

https://doi.org/10.1371/journal.pone.0196878.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Responses by degree.

https://doi.org/10.1371/journal.pone.0196878.t001

No difference by gender or minority-serving institution status.

Analyzing the results for the fifteen survey skills, responses did not vary based on gender or whether the respondent was at a minority-serving institution (see Materials and Methods and [34]). Respondents were overwhelmingly white and non-Hispanic/non-Latino (Fig 2). Individuals who identified as people of color (POC) and those who identified as Hispanic/Latino consistently rated the importance of every skill higher (and often significantly higher, i.e., indicated that it was more important) than those who did not identify as POC or Hispanic/Latino [34]. However, given the tremendous disparity in sample sizes (Fig 2), and known cultural differences in responses to surveys [38, 39], we hesitate to assign much weight to these particular differences.

Higher ratings of some skills at larger institutions.

Respondents gave some survey skills different scores depending on the size of their institution, whether that is determined by total number of students, total number of undergraduates, or number of faculty in the department. Respondents at institutions with fewer than 5,000 total students or 5,000 undergraduates rated the importance of S1 (Role), S2 (Concepts), S13 (Scripting), and S15 (Computational environment) significantly lower (i.e., indicated that they were less important) than those at institutions with more than 15,000 total students or 15,000 undergraduates [34]. Similarly, faculty in larger departments gave significantly higher ratings to S1 (Role) and S2 (Concepts) than those in smaller departments (S1: P = 6.73 × 10^-3 for a two-tailed KS test; S2: P = 1.49 × 10⁻³).

Among the four Carnegie Classifications, there were no significant differences in Likert ratings between respondents at Baccalaureate and Master’s institutions (S1 Table). In contrast, respondents at Associate’s institutions routinely rated every skill lower than did those at other institution types, whereas those at Doctoral institutions rated every skill higher than those at institutions with other classifications. In general, the rating of S13 (Scripting) increased with Carnegie Classification (Fig 4 and S1 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Mean Likert responses for S3 (Statistics) and S13 (Scripting).

Mean Likert responses are shown for (A) S3 (Statistics) and (B) S13 (Scripting) for three categories: Carnegie (Carnegie Classification of the respondent’s home institution: Associate’s, Baccalaureate, Master’s, Doctoral), Year Earned (year that the highest degree was earned; responses were grouped in the following bins: Before 1980, 1980 to 1989, 1990 to 1999, 2000 to 2009, and After 2009), and Training (level of bioinformatics training: None, Self-taught, Short workshop, Undergraduate/PostBacc training, Graduate class, and Graduate degree). Means and P values from pairwise KS tests are reported in [34].

https://doi.org/10.1371/journal.pone.0196878.g004

Higher ratings of categories by year and level of bioinformatics training.

The ratings of most of the fifteen survey skills did not depend significantly on the year in which the respondents earned their degree. However, skills S3 (Statistics) and S13 (Scripting) were rated significantly higher by respondents who earned their degree after 2009 (i.e., in 2010 or after) than by those who earned their degree prior to 2010 (Fig 4 and [34]). Furthermore, the majority of respondents had a PhD, and those with a PhD consistently rated every skill higher than those with a master’s degree, often significantly higher (Table 1). Those with no training in bioinformatics routinely rated every skill significantly lower than respondents with training, formal or not (Fig 4 and [34]). Notably, S13 (Scripting) was rated significantly higher by respondents with a graduate degree in bioinformatics than by those with any other type of training (Fig 4 and [34]).

Coverage of skills varies across syllabi

To analyze the syllabi submitted by survey respondents (see Materials and Methods), the fifteen survey skills were divided into two groups—those that required students to be familiar with a concept (i.e., to “know” about it; survey skills S1 to S4, S6, S8, and S10), and those that required direct engagement (i.e., to be able to use the skill in “practice”; skills S5, S7, S9, and S11 to S15). More of the “knowing” skills were rated as being either “extremely important” or “very important” than the “practice” skills and were more frequently covered in the syllabi; conversely, the “practice” skills were less likely to be covered (Fig 5). There were two exceptions to these trends. Few survey respondents thought students should be expected to be familiar with metabolomic and systems biology data (S10, a “knowing” skill), nor was it frequently covered in the submitted syllabi. (As described below, this skill was dropped from the Core Competencies.) On the other hand, a majority of respondents indicated that undergraduate life scientists should “be able to use bioinformatics tools” (S5). In addition, evidence of this was present in approximately 70% of the submitted syllabi (Fig 5, Fig 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Importance of the fifteen bioinformatics skills as rated by survey respondents compared to coverage of the skills in the syllabi.

Skills are shown with the proportion of survey responses rating the skill as either “Very Important” or “Extremely Important” (blue bars) and the proportion of submitted syllabi that exhibited evidence of the skill (grey bars). Skills requiring familiarity with a concept (“knowing” skills) are to the left of the vertical dashed line; skills requiring direct engagement (“practice” skills) are to the right. In their respective categories, skills are presented in order of decreasing proportion of survey responses rating the skill as Very or Extremely Important.

https://doi.org/10.1371/journal.pone.0196878.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Histogram of number of skills covered per syllabus.

The number of syllabi that addressed the specified number of survey skills (S1 to S15). For example, ten syllabi addressed only one of the fifteen skills. On average, a syllabus covered 5.5 skills, with a median of 6 skills addressed. In aggregate, the submitted syllabi covered all fifteen skills, but no single syllabus covered more than thirteen out of the fifteen skills.

https://doi.org/10.1371/journal.pone.0196878.g006

Syllabi also varied substantially in the number of survey skills covered. On average, a syllabus covered 5.5 skills, with a median of six skills addressed. Although the submitted syllabi, in aggregate, covered all the survey skills, no single syllabus covered more than thirteen out of the fifteen skills, suggesting the difficulty of covering all the skills in one class (Fig 6).

Core competencies

The nine NIBLSE Core Competencies, C1 to C9, developed using the iterative process described in Materials and Methods, are given in Table 2. During the development process, several survey skills that received lower ratings on the survey (S10, S11, S12) were dropped or combined into a more general competency such as C4 (“Use bioinformatics tools to examine complex biological problems”) and C6 (“Explore and/or model biological interactions, networks, and data integration using bioinformatics”). While scripting and use of the command line (S13) received relatively lower scores from survey respondents, the strong support for these emerging skills at Doctoral institutions and from the most recently trained respondents led to their inclusion in the final list. As mentioned above, the NIBLSE Core Competencies are intended to serve as a guide for institutions as they integrate bioinformatics into their own life sciences curricula.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. The NIBLSE bioinformatics core competencies for undergraduate life scientists.

https://doi.org/10.1371/journal.pone.0196878.t002

Discussion

Faculty from a wide range of institutions showed strong support for greater integration of bioinformatics into undergraduate life sciences curricula: 95% of the 1,260 respondents agreed with the statement “I think bioinformatics should be integrated into undergraduate life sciences education.” However, there are some differences among faculty perspectives at different types of institutions. While most respondents value the ability to retrieve information from public databases and use existing software tools to analyze data, respondents at Doctoral institutions place a higher priority on computational skills, such as being able to operate in multiple computational environments and being able to write short programs. A possible explanation for this difference is that respondents at research-based institutions are more directly exposed to the necessity of using these computational skills on a day-to-day basis, whether by themselves, their students, or their colleagues. These findings provide insight into different educational perspectives and the barriers institutions may face as they integrate bioinformatics into their own life sciences curricula.

It is important to keep in mind the distinctions between the educational needs of bioinformaticians and life scientists as well as the differences between the goals of undergraduate education and those of graduate or professional education. Up to this point, much of the discussion of bioinformatics education in the literature has focused on the education of bioinformaticians or on graduate or professional development. However, some authors have addressed the question of bioinformatics education for undergraduate life sciences students. In particular, attendees of the first and second Workshop on Education in Bioinformatics and Computational Biology in 2008 (Taipei, Taiwan) and 2009 (Singapore), held as part of the International Conference in Bioinformatics, attempted to identify a minimum skill set for the training of bioinformaticians and life scientists with informatics capabilities [8]. A consensus list of five essential bioinformatics skills was reported. This skill set overlaps considerably with the NIBLSE Core Competencies, indicating that there is agreement about the skills that are necessary for modern life sciences students. However, the sample size of this effort was small (n = 56, including students) and the authors were attempting to find a consensus that could be applied internationally in widely differing contexts.

The NIBLSE Core Competencies include a computational competency, consistent with other reports [7, 15, 18, 19]. Although many life sciences programs are not currently equipped to provide training in basic programming and operating in a command-line environment, these are important skills that enable students to manipulate and analyze modern biological data. (As an added benefit, creating classes to teach these skills provides an opportunity for interdepartmental course development.) This does not mean that we recommend that life sciences students be able to write complex software applications or be trained to develop graphical user interfaces, but being able to write short programs and run command-line programs gives them flexibility in analyzing data and, perhaps more importantly, provides them with a better understanding of the data itself. The pedagogical literature from a variety of fields is clear that students learn more when they engage with data more deeply, as opposed to entering data into a “black box” and reporting the results [16, 40–44]. Thus, as with any laboratory technique—e.g., PCR, dissection, or microscopy—bioinformatically-literate undergraduates don’t need to be experts but should be expected to have basic skills in these areas when they graduate.

In addition to the survey data, the analysis of syllabi submitted by survey respondents shows that a variety of bioinformatics topics are already being covered in life science curricula nationwide. The fact that bioinformatics can be performed relatively inexpensively with freely available data and software makes it an attractive way for students to engage in research experiences and inquiry-based learning [3, 40, 41]. We would argue that this training would ideally occur in an integrated manner throughout a life sciences curriculum, as opposed to being isolated in a single course. Thus, there is no need to remove particular topics and replace them with “bioinformatics” units. Instead, we encourage faculty to find ways to incorporate bioinformatics techniques and applications as a way of exploring existing concepts in their curricula.

Among the syllabi analyzed, there appears to be a greater emphasis on “knowing” rather than “doing.” This gap may reflect the degree of instructor training: if an instructor has little training or experience with bioinformatics, it is understandably easier to introduce a technique or concept in a lecture than it is to develop and implement an in-depth exercise. This idea is supported by the fact that survey respondents frequently commented on the lack of available teaching resources in this area. One effort to address this problem was the inclusion of bioinformatics as a course on CourseSource, which publishes evidence-based learning resources [33]. The issue of training was also raised repeatedly in responses to survey questions regarding real and perceived barriers to the integration of bioinformatics into life sciences education, with many respondents indicating that they wanted to integrate more bioinformatics into their courses but felt they lacked the necessary training to do so. This training deficit is a long-term problem that will be difficult to address. On the one hand, a variety of resources are currently available for faculty to receive training in bioinformatics or to educate themselves to address the training need. On the other, many faculty in the survey indicated they needed more training. Thus, it’s not clear if the lack of training is due to faculty not being aware of the training resources available or if the existing resources are not useful to them because they cost too much, require too much time, are too advanced, or require too much work to adapt to their specific courses. Barriers to integrating bioinformatics into life sciences education, and our suggestions for overcoming them, are the topics of a separate paper [32]. Note that NIBLSE was formed in large part to provide resources to faculty to help overcome these kinds of barriers.

To conclude, the analysis presented here provides evidence of strong, widespread agreement that undergraduate life sciences students need to be trained in bioinformatics as well as considerable agreement about which skills are necessary. Furthermore, it resulted in development of the NIBLSE Core Competencies, which provide a framework of topics for this training. Although the results presented here could potentially be skewed—those interested in integrating bioinformatics into biology education could have been more motivated to respond to the survey—given the large number of responses [45] and the small margin of error in them (see Materials and Methods), the effects of such a skew, if present, would be minor. The minimal effects of a potential skew are further supported by the close alignment of our results and those in foundational works cited earlier [cf. 8, 9, 27, 28]. As such, we contend that our results accurately represent the opinions of the U.S. life sciences education community as a whole.

Supporting information

Acknowledgments

The authors thank the members of the Genomics Education Partnership, Genome Solver, GCAT-SEEK, and NIBLSE networks for the feedback they provided during the development of the NIBLSE Core Competencies. The authors also thank Sarah Moulton of Moulton Editorial Services (Omaha, Nebraska) for editing and formatting the final version of the manuscript.

The Core Competencies Team (M.A.W.S., C.H., M.S., and M.A.P.), analyzed the survey results and syllabi, came up with recommendations for the proposed list of final core competencies, and wrote the first draft of the manuscript. The NIBLSE Core Competencies Working Group (M.A.W.S., C.H., M.S., S.R., A.G.R., T.M.S., E.W.T., L.R.W., and M.A.P) developed the survey and helped to distribute it. The original NIBLSE Leadership Team (A.G.R., E.W.T., E.D., W.M., and M.A.P.) organized the NIBLSE conference and decided on the final wording of the competencies. All authors contributed to the development of the NIBLSE Core Competencies and to the manuscript.