Keywords

1 Introduction

The Bioinformatics comprises the use of technology for retrieval, storage, mathematical, physical-chemical and statistical analyses, and interpretation and integration of data to solve biological problems [1]. With the advancement of technology and its increasing ease of use, Bioinformatics computational resources have become popular, especially in the academic sphere. However, for an easier understanding and manipulation of molecular structures accessed by these tools, it is indispensable that they meet Human-Computer Interaction (HCI) criteria. Usability and Ergonomics should be persistent concerns in Bioinformatics systems given the amount of information they need to handle and visualize using several tools with various choices each.

On Bioinformatics, in an educational context, there are difficulties faced by both teachers and students. Teachers may have difficulties related to the selection of available resources, biological databases, and tools to manipulate and analyze data such as DNA and protein sequences and structures. There are a variety of tools [1,2,3,4], with different interfaces and interactions. The selection of features may require extra time to learn new tools as well as to select them and evaluate their use in different teaching contexts. For students in the learning process, these issues are aggravated because many tools use Bioinformatics technical terms, with which students are still unfamiliar or do not have adequate background to understand them. Some tools do not have friendly interfaces when compared to others used on a day-to-day basis [5]. Studies indicate that problems with Bioinformatics tools usage satisfaction are related to high workload associated with access to information [6], interactions via command line, navigation and tasks’ execution [7, 8] and usability limitations [7, 9,10,11,12]. For a better user experience, some evidences point towards measures that can be adopted to call the user’s attention, such as reducing cognitive and perceptive overload, being agile, intuitive, consistent, efficient, accessible and that can be easily learned [7, 8, 12,13,14,15]. Furthermore, Bioinformatics tools are usually developed or designed by scientists who have little or no experience in user interface design issues, or even by software developers, who often have little understanding of the Bioinformatics users’ needs [14, 16]. These, however, are challenges common to interdisciplinary areas of teaching and research. In short, as emphasized in [17], it is not uncommon for Bioinformatics software users to experience a steep learning curve to perform their tasks.

The work presented here is the partial result of an evaluation process that involved satisfaction of use criteria as a determining factor in the choice of computational resources, considering different teaching and learning strategies in Bioinformatics. From the analysis of various proposals for syllabi and tools available, a case study was conducted in the Bioinformatics discipline, part of the curriculum of an undergraduate degree in Biological Sciences at PUCRS University in Brazil. The goal was to list Bioinformatics computational resources, analyze how they are used, and understand the factors that influence the selection of these resources by the teacher. Thus, it was possible to understand attitudes, feelings, preferences and values that were involved in the selection of Bioinformatics resources, and to compare them with usability and ergonomics criteria.

Data collection was conducted through observation, questionnaire, interview, and documental research. The interview, documental research, and observation sought to identify the tools and methodology used in the classroom for teaching Bioinformatics, as well as to participate in the teaching process in the classroom. The questionnaire aimed to extract the reasons behind the teacher’s selection of tools. It included usability and ergonomics questions so the teacher could compare different tools, some already known but not always used by him/her. The evaluation instrument was based on the questionnaires SUPRQFootnote 1, SUSFootnote 2, SUMIFootnote 3, WAMMIFootnote 4, and QUISFootnote 5, which are standards in the scientific community to evaluate HCI criteria, and intended to maximize the identification of interface problems, interaction, and satisfaction of use. We also considered the ergonomic criteria [15, 18,19,20,21]. The instrument contained 74 questions directed at the assessment of biological databases (40 questions) and sequences and macromolecule structures viewers (34 questions). These are the tools evaluated:

  • Biological databases: Protein Data Bank Japan (PDBj) [22], Protein Data Bank Europe (PDBe) [23], and The Protein Data Bank (RCSB PDB) [24].

  • Sequences and macromolecular structures viewers: Swiss-PdbViewer (Swiss) [25], VMD - Visual Molecular Dynamics [26], and PyMOL - Molecular Graphics System [27].

The final work discusses reflections on the use of technological resources to support the teaching of Bioinformatics, assessment on the satisfaction of use, usability and ergonomics, aimed at Bioinformatics tools, as well as a discussion of the adopted assessment procedures and the results of the data gathering. The article also presents recommendations for interface design and interaction for Bioinformatics tools.

2 Background

Despite the growth in the area of Bioinformatics since the mid-1950s, due to the exponential increase in computer data generation and processing capacity and advances in the area by the union of computational and experimental information on biological macromolecules [28], it was only in 2001 that Education in Bioinformatics begins to be more prominent. The International Society for Computational Biology (ISCB) has organized the Workshop on Education in Bioinformatics (WEB), which deals with educational and pedagogical issues to determine the nature, extent and content, and the tools available for undergraduate and training programs in Bioinformatics [29]. However, only in 2005 did Bioinformatics education receive more formal recognition when, for the first time, the ISCB includes Bioinformatics Education as a session at the conference [29].

The ISCB, through the Education Committee (EduComm), promotes education and training in Computational Biology and Bioinformatics, offering resources and advice to organizations interested in the development of educational programs [30, 31]. EduComm was responsible for organizing a pilot project that resulted in the creation of a core curriculum for Bioinformatics, called the Task Force Curriculum [30, 32]. In 2012, [30] presented a set of curricular guidelines for education in Bioinformatics. From these, the Global Organization of Bioinformatics Learning, Education and Training (GOBLET) was created, which aims to provide global and sustainable support in Bioinformatics for teachers and apprentices [33].

Some papers deal with perspectives and challenges in the Bioinformatics area [9, 29, 34,35,36,37], and its interdisciplinary nature [31, 38]. There are also different systems that can be used to support the teaching of Bioinformatics. In order to organize these features in terms of use and functionality, the classifications of [39, 40] were used.

Portals such as ExPASyFootnote 6 and the National Center for Biotechnology InformationFootnote 7 (NCBI) are online resources that provide access to a variety of sources of information (biological databases and computational tools) in different areas of Bioinformatics (proteomics, genomics, transcriptomics, phylogeny, others). Biological databases, such as GenBank [41] and RCSB PDB, for example, involve both storage and methods of maintenance, extraction and visualization of information [42] for nucleotide, amino acid sequences or three-dimensional (3D) protein structures. There are also online applications and tools for sequence analysis, such as FASTA [43], BLAST [44], ClustalW [45], which involves sequences of macromolecules (nucleic acids and proteins) and perform alignment of these sequences to identify similar regions between biological sequences. Applications such as Swiss-PdbViewer and PyMOL, are sequence and macromolecule structure viewers, which encode atomic coordinates of PDB data files [46] into three-dimensional (3D) structures, obtained from biological databases. There are also offline tools for analysis of 3D structures, which allow to verify the chemical accuracy of a protein structure, such as PROCHECK [47], and online tools, applications and simulators for prediction of 3D protein structures, such as Modeller [48], I-TASSER [49], which generate 3D structures from a sequence of amino acids.

In view of this large number of tools, it is important to evaluate HCI criteria to alleviate possible usability and ergonomic barriers, which are obstacles in the use of interactive systems. This question becomes more complex when there is a wider range of user profiles such as in Bioinformatics area.

With this orientation, some researches have evaluated the usability in Bioinformatics tools using different methods. The papers of [5,6,7, 17, 50, 51] describe end-user evaluation while papers like [7, 12, 17, 52] describe the use of inspection evaluation. In this way, the objectives of the evaluations were varied. There were studies that analyzed the criteria of efficiency, usability and user preferences regarding the navigation control in tools for navigation in genetic sequences [17], evaluated layout and navigation [51], ease of use [5, 6]. Other studies proposed user-centered design criteria, considering both novice and experienced users [5] and another adapted the Nielsen heuristics [53] to evaluate Bioinformatics tools [52]. We used different tools such as BLAST [6, 17], Varsifter [51], KGGSeq [51], CATH [7], BioCarta [7], Swiss-Prot [7], NCBI [5, 7, 17], Ensembl [17], MEGA [17], among others.

There are also researches that apply user-centered design techniques (personas, interviews, card sorting, usability testing, etc.) in the development of software for the area of Bioinformatics [54, 55]. Overall, these papers argue that the Bioinformatics community knows that there are many resources to access information and visualize it. However, they also agree that there are many interface problems that make it difficult to use such systems.

3 Case Study: Methodological Course

Considering different proposals of Bioinformatics curricula and the experience in the use of tools available for teaching contents of this area, the question of this research was related to the reasons that make a teacher select certain tools to use in the classroom. To that end, the focus of this study was the discipline of Bioinformatics, which integrates the curricular structure of the undergraduate course in Biological Sciences of the Pontifical Catholic University of Rio Grande do Sul (PUCRS). It has a weekly workload of 4 h and a total of 60 classroom hours, being offered in the 4th period of said course. The discipline of Molecular Biology is a prerequisite for this one.

The discipline’s objective is to give the student conditions to know and understand the Bioinformatics main concepts, how to use computational resources for visualization and manipulation of sequences and macromolecular structures, how to locate and use the main Bioinformatics tools available on the Internet, to recognize important problems in the area of Bioinformatics, being able to discuss them with scientists from other areas, and to develop a perception about Bioinformatics literature.

3.1 Subject Matter Expert (SME)

The SME has 26 years of experience in Bioinformatics research, working for 15 years as a Professor in the area both in undergraduate (G) courses in Biological Sciences, Computer Science, Nursing and Pharmacy as in post-graduate (PG) courses in Cellular and Molecular Biology, Computer Science, Science and Mathematics Education and in a Master’s degree in Pharmaceutical Biotechnology. The subjects taught were Bioinformatics (G, PG), Programming for Biological Sciences (G), Scientific Methodology (G), Information Technology Applied to Biological Sciences (G), Informatics for Pharmacy (G), Structural Bioinformatics, Introduction to Bioinformatics (PG), Design of Drugs and Molecular Modeling (PG), Comparative Bioinformatics (PG), Special Topics in Bioinformatics and Computational Modeling (PG), among others.

3.2 Data Gathering Process

To had better understand on what guides the teacher’s choices, data were gathered through different instruments. The study was as follows:

  • Preparation I: analysis of the discipline plan, preparation of the interview script and logbook structure.

  • Data Gathering I: interviews were conducted to identify the Bioinformatics tools used in class, their usage and what guided the choice of the teacher to use them. It also occurred by the researcher participating in classes as an observer, registering entries in a logbook.

  • Interpretation I: transcription of interviews and data recorded in the diary. As an outcome, it was chosen to include, in this study, biological databases and visualizers of sequences and macromolecular structures.

  • Preparation II: an online questionnaire was developed with criteria for usability and satisfaction of use. The instrument was based on the SUPRQ, SUS (System Usability Scale), SUMI (Software Usability Measurement Inventory), WAMMI questionnaire and QUIS (Questionnaire for User Interface Satisfaction) questionnaires. The final version of the questionnaire contained 40 questions applicable to biological databases: 35 closed (using Likert scale) and 5 open. There were 2 questions of identification, 21 of usability, 6 of satisfaction of use and 11 related to use as a resource for teaching and learning process. For sequence and macromolecule structures viewers, there were 34 questions (31 closed and 3 open), of which 2 were of identification, 14 of usability, 6 of satisfaction of use and 12 related to learning.

  • Data Gathering II: the professor answered the questionnaire online and did not present doubts in the interpretation of the questions.

  • Interpretation II: the results were discussed and compared with an analysis of the tools according to HCI criteria: usability, satisfaction of use and ergonomics.

4 Results

4.1 Data Gathering I

The data gathering was carried out to indicate what is taught in the case study discipline and the tools that are used in the classroom, for later identification of the reason for choosing these tools based on HCI criteria.

The discipline content is split in units, per the syllabus: I - Molecular Biology and Informatics; II - Visualization and Computational Manipulation of Three-Dimensional Structures of Biological Macromolecules; III - Pairwise Biological Sequence Alignments; IV - Methods for Multiple Alignments of Biological Sequences; V - Computational Methods for the Prediction of the Secondary and Tertiary Structure of Proteins.

The textbooks used are [39, 40, 56]. Access to the genome and protein data files, tools that have been developed to work with these files and types of questions that these data and tools can answer can be seen in [40]. The reference [39] has a more focused approach on tools that are used in Bioinformatics and [56] provides a critical and comprehensive analysis of the computational methods necessary for the analysis of DNA, RNA and protein data, with basic explanations about the Algorithms, computational methods and strategies for their application to biological problems. The students are also exposed to scientific papers from recent publications of database [41], tools [57], computational methods [58] and special volumes of Nucleic Acids ResearchFootnote 8 , Footnote 9 , Footnote 10.

It was observed the methodology used in classes is theoretical and practical, since students need to know the concepts necessary to solve biological problems and learn to use the computational resources in practice effectively. In theoretical classes, biological contents are presented together with the computational resources used to manipulate information, in a systematic way, linking content and the tool used in practice. The criteria used by the professor to select the resources adopted are:

  • Self-contained tools: it tries to identify the most complete tools, that do not have dependencies of other tools in the execution of a specific functionality.

  • Origin of tools: seeks to value the origin of knowledge, using tools well established in the scientific and academic environment, that are authoritative and keep up to date and with good performance.

  • Ease of learning: the main features of the tools should be clear and easily accessible, so that students can assimilate their content and use it quickly.

  • Tool updates: because Bioinformatics tools are updated frequently, teachers and students need to be aware of content and interface updates.

  • Offline and online tools: offline tools must be accessible to students (free software) and easy to install. As for sequencing tools and biological databases, online tools are preferred, for the convenience of access and the most up-to-date information.

  • Integrated tools: tools that integrate and are validated by others, so that the output of one tool can serve as input to another are preferred. This clearly demonstrates how the workflow of a Bioinformatics researcher looks like.

  • Content filtering: because of the large amount of information it is important that the tools have filters to customize their searches and preferences.

  • Interface standards: familiar to users, like other classic interfaces, such as word processors, for greater ease of use and memory recall.

  • Operating system: applications with Windows interfaces are preferred, for their ease of use, opposed to Linux command-line interfaces, common in these tools.

Considering these criteria and his pedagogical experience as a teacher, he mentioned that he uses: portals - NCBI and ExPASy; biological databases - GenBank, RCSB PDB, CATH - Class, Architecture, Topology, Homology [59] and SCOPe - Structural Classification of Proteins – extended [60]; sequence and macromolecule structures viewers - Swiss-PdbViewer (Swiss); tools for sequence analysis - BLAST and ClustalW; tools for predicting 3D structures - Swiss-Model [61] and Modeller. In addition to all these tools, he uses the educational portal of RCSB PDB, the PDB-101 [24].

From the tools above mentioned, two types are used in more than one learning unit, complementing the use of other tools, constantly participating in theoretical/practical classes and are essential, according to the professor, for the progress of the discipline: biological databases and sequence and macromolecular structures viewers.

During the discipline, the professor uses the RCSB PDB database, which was compared to PDBe and PDBj databases, chosen for having the same type of deposited data, in equal quantity, but that are presented differently in their interfaces. He also adopts the Swiss viewer, compared to VMD and PyMOL, chosen for having interfaces with similar features for the visualization of the structures, to be free and accessed offline. For all the tools, the teacher was asked to examine their interface and answer the questionnaire, the results of which are described in the following subsection.

4.2 Data Gathering II - Results - Part 1: Biological Databases

The professor has already used all the RCSB databases: PDB, PDBe and PDBj, having more experience in RCSB PDB. On his view, all three databases meet usability criteria such as ease of use, ease of navigation, user help and efficiency, and fast response to queries. Icons are intuitive and menus are organized properly.

He further agrees the three databases have a logical structure for biological databases, with easily identified search fields, search filters, and the possibility of querying from different input fields. He also pointed out that the functionalities of the three databases are well integrated. It considers the usefulness of the information to be adequate, with easily understandable and clearly demonstrated results. Information is reliable and up to date. When it comes to satisfaction of use, he agrees that he are comfortable using them, that they look simple and clean, and that he can customize the information that is displayed. However, he stated that PDBj interface could be improved because it is very textual. With the exception of the eight (8) questions listed in Table 1, the feedback provided was very positive.

Table 1. Questions with mixed feedback for RCSB PDB, PDBe and PDBj
Full size table

The professor uses RCSB PDB as an educational tool and as a researcher. He mentioned he uses this database because data files are updated weekly, its interface receives constant improvements and does not need to perform error handling. He strongly recommends this database for teaching Bioinformatics, but has a neutral opinion when comes to PDBe and would not recommend PDBj. In his opinion, the greatest difficulty in using these databases in the classroom is the English language, universal for scientific research, but not accessible to users who do not understand the language.

4.3 Data Gathering II - Results - Part 2: Sequences and Macromolecular Structures Viewers

The Swiss and VMD viewers have been used by the professor since 1998, and PyMOL since 2006. Swiss is adopted in the classroom and as a researcher, at the frequency of more than once a week. The others are used in scientific research once a week.

In the usability criteria, it has evaluated that the three viewers have simple and clean appearance, help options to understand the functionalities and allow the user to control the updates of the tool. He also said that he feels completely comfortable in visualizing a protein in Swiss and PyMol, and that they present information about the structure in a clear and comprehensible way.

He agrees that Swiss-PdbViewer is easy to use and has useful interface components, intuitive icons, and well-integrated features. The other viewers partially meet these features. The preference for Swiss as a viewer can also be noted in Table 2, which brings the issues that have had a negative and neutral feedback.

Table 2. Questions with neutral and negative feedback
Full size table

In this sense, he considers that Swiss is a sophisticated, free, and easy-to-learn viewer and manipulator. He also mentioned criteria such as ease of use, portability, and speed in handling structure files. However, he says its interface is partially attractive. He recommends its use as a teaching resource for Bioinformatics, as well as PyMOL to create images for published papers, due to the good quality of rendered visualizations and generated images.

In his opinion improvements could be made to each of the viewers. Swiss could undergo a graphical rendering of different types of protein structure representation, the VMD menu might be clearer and have easier access options, and the suggestion for PyMOL would be a control panel similar to Swiss.

4.4 Interpretation II

The analysis performed by the professor on the databases and viewers was compared to ergonomics and usability criteria, to verify the reasoning behind the professor’s choice. The criteria were based on - article [18]; Set of criteria for interface ergonomics and user experience of Cybis [15], based on Nielsen’s Heuristics, Shneiderman Gold Rules, Android Design Principles, Principles of ISO/ABNT 92: 41: 110 and Bastien and Scapin ergonomic criteria.

These criteria were translated into recommendations for design and interaction for biological databases and sequence and protein viewers, which are desirable in a good interface and therefore represent principles of good practice that can be adopted by the Bioinformatics interactive systems. The criteria used for evaluation are identified by number, which are followed in subsequent discussions.

Criteria for biological databases

  • CBD1 - Interaction experience: The interface should be pleasant and beautiful for the user, motivating a good experience of use, without forgetting to consider aspects such as effectiveness and efficiency to carry out the tasks.

  • CBD2 - Good use of the screen: The database must have a logical organization to access information, to simplify access and to aid learning. Components with similar features should be close and non-related should be further away without overloading the interface with many elements. The groups can be distributed according to the actions that they apply, for example, in chronological order or importance of the task.

  • CBD3 - Ease of access to information: Key information must be in evidence, clearly available and easily retrieved. The user should not perform multiple interactions to access them. It should be clear where the search field is, search filters, information retrieved at the database, external resources, database statistics, and help and documentation.

  • CBD4 - Visibility of the information and of the system: The database must keep the user informed about the location that is in the interface of the database. Input information and results must be visible, clear, and legible. Buttons and shortcuts of user interest (download, save and refine the search) need to be visible and close to the query field. You should also flag the tools built into the database.

  • CBD5 - Consistency and Interface Standards: Users feel more comfortable and are more efficient if database features maintain a pattern throughout the interface. Menus, search components, filters, query result fields must have the same color pattern, location, and configuration, to facilitate the identification of the resources in the interaction and the user’s learning, so that the user can use it quickly.

  • CBD6 - Data entry made easy: The search field for data search should preferably be at the top of the screen and highlighted, the standard commonly adopted in biological databases. This field should differentiate itself from other search fields, such as from search engines on the internet, for example. In addition, the user must be informed about possible input types, without any ambiguities or doubts. It should have the possibility of refinement of the search (advanced search), and clearly present help and examples of submissions.

  • CBD7- Clarity of information on components and functionalities: The components should be related to their use, without any ambiguities or doubts as to their functions. Metaphors must be clear and easily understood.

  • CBD8 - Accessibility of information: The database must be prepared for assistive technology resources, such as screen readers, magnifiers, auto-contrast, to facilitate access to users who have some type of information access barrier. Accessibility design patterns in database design, such as description of the images and shortcuts by keyboard commands should be observed.

  • CBD9 - Database resource adaptability: The database must be prepared to serve different user profiles, be flexible so tasks can be performed in different ways, and allow customization of the interface for less experienced users. The interface may not display complex information, embedded software, links, and other resources.

  • CBD10 - The Database User Experience: The database should provide the means for experienced users to perform tasks more quickly, through shortcuts and tools that can assist in their work. Similarly, for non-expert users, systematic demonstrations, quick help buttons, input field samples, and data download types, should be made available.

  • CBD11 - Legibility: Because a database has a lot of textual information, it must be legible so that it is accessible to different types of users, such as elderly people, for example. Features to facilitate reading, such as font size, letter/ background contrast, line spacing, line length, should be respected.

  • CBD12 - Decreased workload: This criterion advocates that the user should be spared cognitively and perceptually. It should save you having to perform repeated actions, such as database entries, and save the user unnecessary readings of information. Inputs and outputs must be concise, providing adequate default values.

  • CBD13 - Information Density: The density of information displayed on the screen must be taken care of, so that beginning users do not have difficulty in finding what they need in the database interface. In this way, we recommend that only items related to the user search are presented, and there may be information filters, hidden panels, and menus for accessing other information.

  • CBD14 - Updates to the database: The databases must clearly display in the interface to the user data about their updates, such as the date of the last update, the amount of data deposited, and how long the database is updated.

  • CBD15 - Ease of access to results: The results of a query should be easily accessible, objective, and clear. They should be arranged in a prominent place on the interface and allow customization with filters. Tools that facilitate access to these results can be made available, such as the adequacy of tools to visualize these data directly in the interface and tools for the analysis of this data.

  • CBD16 - Management, protection, and prevention of errors: The database must have a secure interface to the user. You should be prepared to detect and prevent data entry or command errors. User input can be suggested from the character typing.

  • CBD17 - Tolerance to errors: A database with tolerance minimizes risks and negative consequences due to accidental or involuntary actions.

In the evaluation of the database, a search of information was carried out and the initial pages, the search field, the result page of the query, the page displaying the selected result, statistics and documentation were checked. Additional tools such as visualizers, among others, were also observed. The three databases met CBD5, CBD10 and CBD15 criteria. Table 3 shows the criteria that the databases did not meet or partially met, along with the justification.

Table 3. Evaluation of ergonomic and usability criteria for biological databases.
Full size table

Criteria for sequence and macromolecular structures viewers

  • CV1 - Interaction Experience: it is related to the interface to have a good aesthetic, but keeping the application working properly, for an effective and pleasant experience. In addition, the viewer should be simple to the point where the user does not require intensive learning and read extensive documentation before using.

  • CV2 - Message and Functionality Adequacy: The viewer functionality should provide instructions for its use. These should be clear, objective, and easy to understand. The language should be simple and speak the language of the user. Technical terms should be avoided. However, when necessary add quick help buttons and active help, with explanation of terms and suggestions of actions for the user.

  • CV3 - Immediate feedback of actions: When the user performs an action in the viewer the result should be presented quickly and be easy to be noticed. If the action is in progress, the user must be informed that the processing is being performed. Progress indicators should be visible and have the option to cancel, or allow another activity to be performed during its execution.

  • CV4 - Functions with easy access and grouped by type of interaction: The interface must be functional, keeping key tasks at a rapid range and grouped by the type of data they manipulate through.

  • CV5 - Viewer’s Feature Adaptability: The viewer should allow its functions and screens to be customizable, flexible, and adaptable to inexperienced and experienced users. Different paths can be made available for a feature that is accessed frequently, such as the same function in an icon in the toolbar, in the menu, and by keyboard shortcut.

  • CV6 - Error handling, protection, and prevention: This criterion recommends that the viewer interface must be safe for the user to manipulate their protein without worrying about losing the changes made or overwriting manipulated PDB files without due warning.

  • CV7 - Interface homogeneity: the interface must respect the standards and styles of the operating system platform to which it was developed. Consistency is important because a novice user can repeat previously used strategies in interaction with other known applications. Commonly used commands, menus, and icons should be easily recognized, located, and stable on the same screen location. This facilitates learning because the system becomes predictable, intuitive, and more easily remembered.

  • CV8 - Decreased workload: The user should be spared from cognitive and perceptive overload, simplifying the actions to be performed in the viewer. The interaction should be as simple as possible, avoiding to require many actions to complete it. For manipulation of a macromolecule, the commands should be minimized to the maximum, avoiding that the user needs to go through several menus or screens, or having to perform actions through the command line.

  • CV9 - User control and freedom: The user must be able to control the interactions with the viewer, having commands to undo and redo a certain task, control decreases the probability of errors and favors learning. This is important for the user to explore the interface by applying different commands on the protein and verifying the actions they perform.

  • CV10 - Help and documentation: Refers to user support in learning and helping you to do an activity. It can present the user with help in contextual, conceptual, systematic, tutorial style and quick help buttons. The application should also bring informational and technical documentation, which assist users with possible doubts.

The three viewers analyzed have interfaces that are similar, with a large viewing area and additional windows with some commands and main menu. These components were analyzed and the results are shown in Table 4.

Table 4. Evaluation of ergonomic and usability criteria for sequence and macromolecular biological viewers.
Full size table

Based on what was presented in the results it was observed that it is possible to establish relations between the software choices used by the professor and ergonomic and usability criteria. The following are some of the most obvious relationships.

According to the professor’s assessment, Swiss did not have any negative feedback, while for VMD it was registered 10 negative questions (related to usability, satisfaction of use and learning) and to PyMOL, 6 negative questions (related to menu organization and memorizing commands). He did not know how to comment on the error handling criteria and interface customization in VMD’s case.

In the evaluation performed according to 10 HCI criteria for viewers, the CV3 criterion, immediate feedback of the actions, was the only one that was attended in equality in the three applications. Swiss assessment was positive, partially meeting the CV2 and CV6 criteria and in full to the others. This is in agreement with the professor’s positive assessment. VMD met only two criteria: management, protection and prevention of errors and immediate feedback of actions, which may suggest why the professor’s negative evaluation. PyMOL obtained 5 positive criteria: CV3, CV4, CV5, CV8 and CV9, presenting only one negative barrier regarding interface homogeneity - CV7.

In the evaluation of the 17 criteria of HCI for biological database all the database attended CBD5, CBD10 and CBD15 criteria. RCSB PDB met 13 criteria and 3 partially, PDBe met 8 criteria and 8 partially and PDBj met 6 criteria and 8 partially. Criterion CBD8, accessibility of information, was not met in all three databases. It should be remembered that this evaluation was automatic, being recommended a manual revision, essential to prove the accessibility of the page, verification that will be dealt with in future works. PDBj also had a negative assessment of readability and decreased workload (CBD11, CBD12), and the professor also pointed out the excess of textual information in an open question related to interface improvements.

Despite these differences, the professor evaluated most of the database instrument issues positively. The downside of his assessment is that database users need technical knowledge to be able to consult the database, which can make access difficult for non-expert users. He adopts the RSCB PDB by its weekly updates for interface improvements, and does not require error handling. He has also recognized the authority of the database, since RCSB PDB was the first primary database of 3D structures for biological macromolecules.

5 Conclusion

This work is part of a doctoral research that aims to contribute knowledge about the use of tools in teaching Bioinformatics, based on the analysis of the practical experiences of a teacher. The objective of the doctoral study is to propose and validate a teaching method and analyze how we can represent specific information in the field of Bioinformatics and molecular biology, targeting people with visual impairment. It also proposes to observe which characteristics should be contemplated in the development of accessible/ inclusive interfaces for teaching in this area.

In this part of the study, we sought to analyze the perceptions of a university professor about the teaching-learning process with the use of services and tools in Bioinformatics, providing a set of HCI criteria that can be used to evaluate usability, satisfaction of use and ergonomics of biological databases and visualizers of sequences and biological macromolecules, and which translate into recommendations for interface design and interaction for applications in the Bioinformatics field.

For future work, we suggest extending the case study, carried out in the discipline of bioinformatics of PUCRS, to other institutions and compare the tools used to the criteria described in this study. In addition, the results of the evaluations were measured qualitatively only. We could improve these results with a quantitative evaluation, where we could measure usability, satisfaction of use, and ergonomics of these applications in a less subjective way.