Search

Software development has become an integral part of the scholarly ecosystem, spanning all fields and disciplines. To support the sharing and creation of knowledge in line with open science principles, and particularly to enable the reproducibility of research results, it is crucial to make the source code of research software available, allowing for modification, reuse, and distribution. Recognizing the significance of open-source software contributions in academia, the second French Plan for Open Science, announced by the Minister of Higher Education and Research in 2021, introduced a National Award to promote open-source research software. This award serves multiple objectives: firstly, to highlight the software projects and teams that have devoted time and effort to develop outstanding research software, sometimes for decades, and often with little recognition; secondly, to draw attention to the importance of software as a valuable research output and to inspire new generations of researchers to follow and learn from these examples. We present here an in-depth analysis of the design and implementation of this unique initiative. As a national award established explicitly to foster Open Science practices by the French Minister of Research, it faced the intricate challenge of fairly evaluating open research software across all fields, striving for inclusivity across domains, applications, and participants. We provide a comprehensive report on the results of the first edition, which received 129 high-quality submissions. Additionally, we emphasize the impact of this initiative on the open science landscape, promoting software as a valuable research outcome, on par with publications.

Musical co-creativity aims at making humans and computers collaborate to compose music. As an MIR team in computational musicology, we experimented with co-creativity when writing our entry to the “AI Song Contest 2020”. Artificial intelligence was used to generate the song’s structure, harmony, lyrics, and hook melody independently and as a basis for human composition. It was a challenge from both the creative and the technical point of view: in a very short time-frame, the team had to adapt its own simple models, or experiment with existing ones, to a related yet still unfamiliar task, music generation through AI. The song we propose is called “I Keep Counting”. We openly detail the process of songwriting, arrangement, and production. This experience raised many questions on the relationship between creativity and machine, both in music analysis and generation, and on the role AI could play to assist a composer in their work. We experimented with 'AI as automation', mechanizing some parts of the composition, and especially 'AI as suggestion' to foster the composer’s creativity, thanks to surprising lyrics, uncommon successions of sections and unexpected chord progressions. Working with this material was thus a stimulus for human creativity.

Automatic harmonic analysis has been an enduring focus of the MIR community, and has enjoyed a particularly vigorous revival of interest in the machine-learning age. We focus here on the specific case of Roman numeral analysis which, by virtue of requiring key/functional information in addition to chords, may be viewed as an acutely challenging use case. We report on three main developments. First, we provide a new meta-corpus bringing together all existing Roman numeral analysis datasets; this offers greater scale and diversity, not only of the music represented, but also of human analytical viewpoints. Second, we examine best practices in the encoding of pitch, time, and harmony for machine learning tasks. The main contribution here is the introduction of full pitch spelling to such a system, an absolute must for the comprehensive study of musical harmony. Third, we devised and tested several neural network architectures and compared their relative accuracy. In the best-performing of these models, convolutional layers gather the local information needed to analyse the chord at a given moment while a recurrent part learns longer-range harmonic progressions. Altogether, our best representation and architecture produce a small but significant improvement on overall accuracy while simultaneously integrating full pitch spelling. This enables the system to retain important information from the musical sources and provide more meaningful predictions for any new input.

The musical analysis of large-scale structures, such as the classical sonata form, requires to integrate multiple analyses of local musical events into a global coherent analysis. Modelling large-scale structures is still a challenging task for the research community. It includes building large and accurate annotated corpora, as well as developing practical and efficient tools in order to visualize the analyses of these corpora. It finally requires the conception of effective and properly evaluated MIR algorithms. We propose a machine learning approach for the sonata form structure on 32 movements from Mozart’s string quartets. We release an open dataset, encoding two reference analyses of these 32 movements, totaling more than 1800 curated annotations, as well as flexible visualizations of these analyses. We discuss the occurrence in this corpus of melodic, harmonic, and rhythmic features induced by pitches, durations, and rests. We investigate whether the presence or the absence of these features can be characteristic of the different sections forming a sonata form. We then compute the emission and transition probabilities of several Hidden Markov Models intended to match the structure of sonata forms at several resolutions. Our results confirm that the sonata form is better identified when the parameters are learned rather than manually set up. These results open perspectives on the computational analysis of musical forms by mixing human knowledge and machine learning from annotated scores.

Background Labels are a way to add some information on a text, such as functional annotations such as genes on a DNA sequences. V(D)J recombinations are DNA recombinations involving two or three short genes in lymphocytes. Sequencing this short region (500 bp or less) produces labeled sequences and brings insight in the lymphocyte repertoire for onco-hematology or immunology studies. Methods We present two indexes for a text with non-overlapping labels. They store the text in a Burrows–Wheeler transform (BWT) and a compressed label sequence in a Wavelet Tree. The label sequence is taken in the order of the text (TL-index) or in the order of the BWT (TLBW-index). Both indexes need a space related to the entropy of the labeled text. Results These indexes allow efficient text–label queries to count and find labeled patterns. The TLBW-index has an overhead on simple label queries but is very efficient on combined pattern–label queries. We implemented the indexes in C++ and compared them against a baseline solution on pseudo-random as well as on V(D)J labeled texts. Discussion New indexes such as the ones we proposed improve the way we index and query labeled texts as, for instance, lymphocyte repertoire for hematological and immunological studies.

[This corrects the article DOI: 10.1371/journal.pone.0166126.].

BackgroundThe B and T lymphocytes are white blood cells playing a key role in the adaptive immunity. A part of their DNA, called the V(D)J recombinations, is specific to each lymphocyte, and enables recognition of specific antigenes. Today, with new sequencing techniques, one can get billions of DNA sequences from these regions. With dedicated Repertoire Sequencing (RepSeq) methods, it is now possible to picture population of lymphocytes, and to monitor more accurately the immune response as well as pathologies such as leukemia.Methods and resultsVidjil is an open-source platform for the interactive analysis of high-throughput sequencing data from lymphocyte recombinations. It contains an algorithm gathering reads into clonotypes according to their V(D)J junctions, a web application made of a sample, experiment and patient database and a visualization for the analysis of clonotypes along the time. Vidjil is implemented in C++, Python and Javascript and licensed under the GPLv3 open-source license. Source code, binaries and a public web server are available at http://www.vidjil.org and at http://bioinfo.lille.inria.fr/vidjil. Using the Vidjil web application consists of four steps: 1. uploading a raw sequence file (typically a FASTQ); 2. running RepSeq analysis software; 3. visualizing the results; 4. annotating the results and saving them for future use. For the end-user, the Vidjil web application needs no specific installation and just requires a connection and a modern web browser. Vidjil is used by labs in hematology or immunology for research and clinical applications.