Domain Adaptation in remote sensing: increasing the portability of land-cover classifiers

Among the types of remote sensing acquisitions, optical images are certainly one of the most widely relied upon data sources for Earth observation. They provide detailed measurements of the electromagnetic radiation reflected or emitted by each pixel in the scene. Through a process termed supervised land-cover classification, this allows to automatically yet accurately distinguish objects at the surface of our planet. In this respect, when producing a land-cover map of the surveyed area, the availability of training examples representative of each thematic class is crucial for the success of the classification procedure. However, in real applications, due to several constraints on the sample collection process, labeled pixels are usually scarce. When analyzing an image for which those key samples are unavailable, a viable solution consists in resorting to the ground truth data of other previously acquired images. This option is attractive but several factors such as atmospheric, ground and acquisition conditions can cause radiometric differences between the images, hindering therefore the transfer of knowledge from one image to another. The goal of this Thesis is to supply remote sensing image analysts with suitable processing techniques to ensure a robust portability of the classification models across different images. The ultimate purpose is to map the land-cover classes over large spatial and temporal extents with minimal ground information. To overcome, or simply quantify, the observed shifts in the statistical distribution of the spectra of the materials, we study four approaches issued from the field of machine learning. First, we propose a strategy to intelligently sample the image of interest to collect the labels only in correspondence of the most useful pixels. This iterative routine is based on a constant evaluation of the pertinence to the new image of the initial training data actually belonging to a different image. Second, an approach to reduce the radiometric differences among the images by projecting the respective pixels in a common new data space is presented. We analyze a kernel-based feature extraction framework suited for such problems, showing that, after this relative normalization, the cross-image generalization abilities of a classifier are highly increased. Third, we test a new data-driven measure of distance between probability distributions to assess the distortions caused by differences in the acquisition geometry affecting series of multi-angle images. Also, we gauge the portability of classification models through the sequences. In both exercises, the efficacy of classic physically- and statistically-based normalization methods is discussed. Finally, we explore a new family of approaches based on sparse representations of the samples to reciprocally convert the data space of two images. The projection function bridging the images allows a synthesis of new pixels with more similar characteristics ultimately facilitating the land-cover mapping across images.

Leaf area of common bean genotypes during early pod filling as related to plant adaptation to limited phosphorus supply

Low phosphorus supply markedly limits leaf growth and genotypes able to maintain adequate leaf area at low P could adapt better to limited-P conditions. This work aimed to investigate the relationship between leaf area production of common bean (Phaseolus vulgaris) genotypes during early pod filling and plant adaptation to limited P supply. Twenty-four genotypes, comprised of the four growth habits in the species and two weedy accessions, were grown at two P level applied to the soil (20 and 80 mg kg-1) in 4 kg pots and harvested at two growth stages (pod setting and early pod filling). High P level markedly increased the leaf number and leaf size (leaf area per leaf), slightly increased specific leaf area but did not affect the net assimilation rate. At low P level most genotypic variation for plant dry mass was associated with leaf size, whereas at high P level this variation was associated primarily with the number of leaves and secondarily with leaf size, specific leaf area playing a minor role at both P level. Determinate bush genotypes presented a smaller leaf area, fewer but larger leaves with higher specific leaf area and lower net assimilation rate. Climbing genotypes showed numerous leaves, smaller and thicker leaves with a higher net assimilation rate. Indeterminate bush and indeterminate prostrate genotypes presented the highest leaf area, achieved through intermediate leaf number, leaf size and specific leaf area. The latter groups were better adapted to limited P. It is concluded that improved growth at low P during early pod filling was associated with common bean genotypes able to maintain leaf expansion through leaves with greater individual leaf area.

CHARACTERIZATION OF NrsZ, A NOVEL SMALL REGULATORY NON-CODING RNA INVOLVED IN THE ADAPTATION OF PSEUDOMONAS AERUGINOSA PAO1

The opportunistic ubiquitous pathogen Pseudomonas aeruginosa strain PAOl is a versatile Gram-negative bacterium that has the extraordinary capacity to colonize a wide diversity of ecological niches and to cause severe and persistent infections in humans. To ensure an optimal coordination of the genes involved in nutrient utilization, this bacterium uses the NtrB/C and/or the CbrA/B two-component systems, to sense nutrients availability and to regulate in consequence the expression of genes involved in their uptake and catabolism. NtrB/C is specialized in nitrogen utilization, while the CbrA/B system is involved in both carbon and nitrogen utilization and both systems activate their target genes expression in concert with the alternative sigma factor RpoN. Moreover, the NtrB/C and CbrA/B two- component systems regulate the secondary metabolism of the bacterium, such as the production of virulence factors. In addition to the fine-tuning transcriptional regulation, P. aeruginosa can rapidly modulate its metabolism using small non-coding regulatory RNAs (sRNAs), which regulate gene expression at the post-transcriptional level by diverse and sophisticated mechanisms and contribute to the fast physiological adaptability of this bacterium. In our search for novel RpoN-dependent sRNAs modulating the nutritional adaptation of P. aeruginosa PAOl, we discovered NrsZ (Nitrogen regulated sRNA), a novel RpoN-dependent sRNA that is induced under nitrogen starvation by the NtrB/C two-component system. NrsZ has a unique architecture, formed of three similar stem-loop structures (SL I, II and II) separated by variant spacer sequences. Moreover, this sRNA is processed in short individual stem-loop molecules, by internal cleavage involving the endoribonuclease RNAse E. Concerning NrsZ functions in P. aeruginosa PAOl, this sRNA was shown to trigger the swarming motility and the rhamnolipid biosurfactants production. This regulation is due to the NrsZ-mediated activation of rhlA expression, a gene encoding for an enzyme essential for swarming motility and rhamnolipids production. Interestingly, the SL I structure of NrsZ ensures its regulatory function on rhlA expression, suggesting that the similar SLs are the functional units of this modular sRNA. However, the regulatory mechanism of action of NrsZ on rhlA expression activation remains unclear and is currently being investigated. Additionally, the NrsZ regulatory network was investigated by a transcriptome analysis, suggesting that numerous genes involved in both primary and secondary metabolism are regulated by this sRNA. To emphasize the importance of NrsZ, we investigated its conservation in other Pseudomonas species and demonstrated that NrsZ is conserved and expressed under nitrogen limitation in Pseudomonas protegens Pf-5, Pseudomonas putida KT2442, Pseudomonas entomophila L48 and Pseudomonas syringae pv. tomato DC3000, strains having different ecological features, suggesting an important role of NrsZ in the adaptation of Pseudomonads to nitrogen starvation. Interestingly the architecture of the different NrsZ homologs is similarly composed by SL structures and variant spacer sequences. However, the number of SL repetitions is not identical, and one to six SLs were predicted on the different NrsZ homologs. Moreover, NrsZ is processed in short molecules in all the strains, similarly to what was previously observed in P. aeruginosa PAOl, and the heterologous expression of the NrsZ homologs restored rhlA expression, swarming motility and rhamnolipids production in the P. aeruginosa NrsZ mutant. In many aspects, NrsZ is an atypical sRNA in the bacterial panorama. To our knowledge, NrsZ is the first described sRNA induced by the NtrB/C. Moreover, its unique modular architecture and its processing in similar short SL molecules suggest that NrsZ belongs to a novel family of bacterial sRNAs. -- L'agent pathogène opportuniste et ubiquitaire Pseudomonas aeruginosa souche PAOl est une bactérie Gram négative versatile ayant l'extraordinaire capacité de coloniser différentes niches écologiques et de causer des infections sévères et persistantes chez l'être humain. Afin d'assurer une coordination optimale des gènes impliqués dans l'utilisation de différents nutriments, cette bactérie se sert de systèmes à deux composants tel que NtrB/C et CbrA/B afin de détecter la disponibilité des ressources nutritives, puis de réguler en conséquence l'expression des gènes impliqués dans leur importation et leur catabolisme. Le système NtrB/C régule l'utilisation des sources d'azote alors que le système CbrA/B est impliqué à la fois dans l'utilisation des sources de carbone et d'azote. Ces deux systèmes activent l'expression de leurs gènes-cibles de concert avec le facteur sigma alternatif RpoN. En outre, NtrB/C et CbrA/B régulent aussi le métabolisme secondaire, contrôlant notamment la production d'importants facteurs de virulence. En plus de toutes ces régulations génétiques fines ayant lieu au niveau transcriptionnel, P. aeruginosa est aussi capable de moduler son métabolisme en se servant de petits ARNs régulateurs non-codants (ARNncs), qui régulent l'expression génétique à un niveau post- transcriptionnel par divers mécanismes sophistiqués et contribuent à rendre particulièrement rapide l'adaptation physiologique de cette bactérie. Au cours de nos recherches sur de nouveaux ARNncs dépendant du facteur sigma RpoN et impliqués dans l'adaptation nutritionnelle de P. aeruginosa PAOl, nous avons découvert NrsZ (Nitrogen regulated sRNA), un ARNnc induit par la cascade NtrB/C-RpoN en condition de carence en azote. NrsZ a une architecture unique, composée de trois structures en tige- boucle (TB I, II et III) hautement similaires et séparées par des « espaceurs » ayant des séquences variables. De plus, cet ARNnc est clivé en petits fragments correspondant au trois molécules en tige-boucle, par un processus de clivage interne impliquant l'endoribonucléase RNase E. Concernant les fonctions de NrsZ chez P. aeruginosa PAOl, cet ARNnc est capable d'induire la motilité de type « swarming » et la production de biosurfactants, nommés rhamnolipides. Cette régulation est due à l'activation par NrsZ de l'expression de rhlA, un gène essentiel pour la motilité de type swarming et pour la production de rhamnolipides. Étonnamment, la structure TB I est capable d'assurer à elle seule la fonction régulatrice de NrsZ sur l'expression de rhlA, suggérant que ces molécules TBs sont les unités fonctionnelles de cet ARNnc modulaire. Cependant, le mécanisme moléculaire par lequel NrsZ active l'expression de rhlA demeure à ce jour incertain et est actuellement à l'étude. En plus, le réseau de régulations médiées par NrsZ a été étudié par une analyse de transcriptome qui a indiqué que de nombreux gènes impliqués dans le métabolisme primaire ou secondaire seraient régulés par NrsZ. Pour accentuer l'importance de NrsZ, nous avons étudié sa conservation dans d'autres espèces de Pseudomonas. Ainsi, nous avons démontré que NrsZ est conservé et exprimé en situation de carence d'azote par les souches Pseudomonas protegens Pf-5, Pseudomonas putida KT2442, Pseudomonas entomophila L48, Pseudomonas syringae pv. tomato DC3000, quatre espèces ayant des caractéristiques écologiques très différentes, suggérant que NrsZ joue un rôle important dans l'adaptation du genre Pseudomonas envers la carence en azote. Chez toutes les souches étudiées, les différents homologues de NrsZ présentent une architecture similaire faite de TBs conservées et d'espaceurs. Cependant, le nombre de TBs n'est pas identique et peut varier de une à six copies selon la souche. Les différentes versions de NrsZ sont clivées en petites molécules dans ces quatre souches, comme il a été observé chez P. aeruginosa PAOl. De plus, l'expression hétérologue des différentes variantes de NrsZ est capable de restaurer l'expression de rhlA, la motilité swarming et la production de rhamnolipides dans une souche de P. aeruginosa dont nrsZ a été inactivé. Par bien des aspects, NrsZ est un ARNnc atypique dans le monde bactérien. À notre connaissance, NrsZ est le premier ARNnc décrit comme étant régulé par le système NtrB/C. De plus, son unique architecture modulaire et son clivage en petites molécules similaires suggèrent que NrsZ appartient à une nouvelle famille d'ARNncs bactériens.

Etude de validation d'une version francophone du Q-Sort d'attachement de Waters et Deane = A validation study of the french adaptation of Waters and Deane's attachment Q-sort

Ce texte relate l'étude de validation d'une adaptation francophone du Q-Sort d'attachement de Waters et Deane (cette adaptation est présentée dans ce même numéro). Le Q-Sort rempli par les parents ne paraît pas tenir ses promesses de méthode alternative à la Situation étrange pour évaluer la qualité de la relation d'attachement. Cependant, lorsqu'il est rempli par un observateur extérieur, il semble mieux refléter la catégorisation des comportements dans la Situation étrange. Au travers du Q-Sort, les parents semblent décrire davantage le tempérament de l'enfant que la qualité de la relation d'attachement.

Was the Classical Tradition Betrayed by J. Ivory's adaptation of E. M. Forster's "Maurice"?

According to Literature and Film studies and from the point of view of the influence of Classical Tradition on Western Culture -Classical Greek Tradition, in this case-, this article is an accurate analysis of the inevitable -to a certain degree- screenwriters betrayals regarding the literary texts that they adapt. However, in spite of being practically inevitable, Dr. Pau Gilabert Barberà indicates which are in his opinion the limits beyond which Ivory/Hesketh-Harvey should have not gone in order not to dilute the Hellenic temper of E. M. Forster's Maurice.

Plant adaptation to temperature and photoperiod

Selostus: Kasvien sopeutuminen lämpötilaan ja päivän pituuteen

Cross-cultural adaptation, reliability, internal consistency and validation of the Spinal Function Sort (SFS) for French- and German-speaking patients with back complaints.

Introduction Functional subjective evaluation through questionnaire is fundamental, but not often realized in patients with back complaints, lacking validated tools. The Spinal Function Sort (SFS) was only validated in English. We aimed to translate, adapt and validate the French (SFS-F) and German (SFS-G) versions of the SFS. Methods Three hundred and forty-four patients, experiencing various back complaints, were recruited in a French (n = 87) and a German-speaking (n = 257) center. Construct validity was estimated via correlations with SF-36 physical and mental scales, pain intensity and hospital anxiety and depression scales (HADS). Scale homogeneities were assessed by Cronbach's α. Test-retest reliability was assessed on 65 additional patients using intraclass correlation (IC). Results For the French and German translations, respectively, α were 0.98 and 0.98; IC 0.98 (95% CI: [0.97; 1.00]) and 0.94 (0.90; 0.98). Correlations with physical functioning were 0.63 (0.48; 0.74) and 0.67 (0.59; 0.73); with physical summary 0.60 (0.44; 0.72) and 0.52 (0.43; 0.61); with pain -0.33 (-0.51; -0.13) and -0.51 (-0.60; -0.42); with mental health -0.08 (-0.29; 0.14) and 0.25 (0.13; 0.36); with mental summary 0.01 (-0.21; 0.23) and 0.28 (0.16; 0.39); with depression -0.26 (-0.45; -0.05) and -0.42 (-0.52; -0.32); with anxiety -0.17 (-0.37; -0.04) and -0.45 (-0.54; -0.35). Conclusions Reliability was excellent for both languages. Convergent validity was good with SF-36 physical scales, moderate with VAS pain. Divergent validity was low with SF-36 mental scales in both translated versions and with HADS for the SFS-F (moderate in SFS-G). Both versions seem to be valid and reliable for evaluating perceived functional capacity in patients with back complaints.

Evolutionary dynamics of molecular markers during local adaptation: a case study in Drosophila subobscura

Background: Natural selection and genetic drift are major forces responsible for temporal genetic changes in populations. Furthermore, these evolutionary forces may interact with each other. Here we study the impact of an ongoing adaptive process at the molecular genetic level by analyzing the temporal genetic changes throughout 40 generations of adaptation to a common laboratory environment. Specifically, genetic variability, population differentiation and demographic structure were compared in two replicated groups of Drosophila subobscura populations recently sampled from different wild sources. Results: We found evidence for a decline in genetic variability through time, along with an increase in genetic differentiation between all populations studied. The observed decline in genetic variability was higher during the first 14 generations of laboratory adaptation. The two groups of replicated populations showed overall similarity in variability patterns. Our results also revealed changing demographic structure of the populations during laboratory evolution, with lower effective population sizes in the early phase of the adaptive process. One of the ten microsatellites analyzed showed a clearly distinct temporal pattern of allele frequency change, suggesting the occurrence of positive selection affecting the region around that particular locus. Conclusion: Genetic drift was responsible for most of the divergence and loss of variability between and within replicates, with most changes occurring during the first generations of laboratory adaptation. We also found evidence suggesting a selective sweep, despite the low number of molecular markers analyzed. Overall, there was a similarity of evolutionary dynamics at the molecular level in our laboratory populations, despite distinct genetic backgrounds and some differences in phenotypic evolution.

Organisational adaptation of multi-agent systems in a peer-to-peer scenario

Organisations in Multi-Agent Systems (MAS) have proven to be successful in regulating agent societies. Nevertheless, changes in agents' behaviour or in the dynamics of the environment may lead to a poor fulfilment of the system's purposes, and so the entire organisation needs to be adapted. In this paper we focus on endowing the organisation with adaptation capabilities, instead of expecting agents to be capable of adapting the organisation by themselves. We regard this organisational adaptation as an assisting service provided by what we call the Assistance Layer. Our generic Two Level Assisted MAS Architecture (2-LAMA) incorporates such a layer. We empirically evaluate this approach by means of an agent-based simulator we have developed for the P2P sharing network domain. This simulator implements 2-LAMA architecture and supports the comparison between different adaptation methods, as well as, with the standard BitTorrent protocol. In particular, we present two alternatives to perform norm adaptation and one method to adapt agents'relationships. The results show improved performance and demonstrate that the cost of introducing an additional layer in charge of the system's adaptation is lower than its benefits.

Skeletal muscle mitochondrial and lipid droplet content assessed with standardized grid sizes for stereology.

Skeletal muscle mitochondrial (Mito) and lipid droplet (Lipid) content are often measured in human translational studies. Stereological point counting allows computing Mito and Lipid volume density (Vd) from micrographs taken with transmission electron microscopes. Former studies are not specific as to the size of individual squares that make up the grids, making reproducibility difficult, particularly when different magnifications are used. Our objective was to determine which size grid would be best at predicting fractional volume efficiently without sacrificing reliability and to test a novel method to reduce sampling bias. Methods: ten subjects underwent vastus lateralis biopsies. Samples were fixed, embedded, and cut longitudinally in ultrathin sections of 60 nm. Twenty micrographs from the intramyofibrillar region were taken per subject at Ã-33,000 magnification. Different grid sizes were superimposed on each micrograph: 1,000 Ã- 1,000 nm, 500 Ã- 500 nm, and 250 Ã- 250 nm. Results: mean Mito and Lipid Vd were not statistically different across grids. Variability was greater when going from 1,000 Ã- 1,000 to 500 Ã- 500 nm grid than from 500 Ã- 500 to 250 Ã- 250 nm grid. Discussion: this study is the first to attempt to standardize grid size while keeping with the conventional stereology principles. This is all in hopes of producing replicable assessments that can be obtained universally across different studies looking at human skeletal muscle mitochondrial and lipid droplet content.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Introduction: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on measurement of blood concentrations. Maintaining concentrations within a target range requires pharmacokinetic and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. In the last decades computer programs have been developed to assist clinicians in this assignment. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Method: Literature and Internet search was performed to identify software. All programs were tested on common personal computer. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software's characteristics. Numbers of drugs handled vary widely and 8 programs offer the ability to the user to add its own drug model. 10 computer programs are able to compute Bayesian dosage adaptation based on a blood concentration (a posteriori adjustment) while 9 are also able to suggest a priori dosage regimen (prior to any blood concentration measurement), based on individual patient covariates, such as age, gender, weight. Among those applying Bayesian analysis, one uses the non-parametric approach. The top 2 software emerging from this benchmark are MwPharm and TCIWorks. Other programs evaluated have also a good potential but are less sophisticated (e.g. in terms of storage or report generation) or less user-friendly.¦Conclusion: Whereas 2 integrated programs are at the top of the ranked listed, such complex tools would possibly not fit all institutions, and each software tool must be regarded with respect to individual needs of hospitals or clinicians. Interest in computing tool to support therapeutic monitoring is still growing. Although developers put efforts into it the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capacity of data storage and report generation.

Benchmarking therapeutic drug monitoring software: A systematic evaluation of available computer tools

Objectives: Therapeutic drug monitoring (TDM) aims at optimizing treatment by individualizing dosage regimen based on blood concentrations measurement. Maintaining concentrations within a target range requires pharmacokinetic (PK) and clinical capabilities. Bayesian calculation represents a gold standard in TDM approach but requires computing assistance. The aim of this benchmarking was to assess and compare computer tools designed to support TDM clinical activities.¦Methods: Literature and Internet were searched to identify software. Each program was scored against a standardized grid covering pharmacokinetic relevance, user-friendliness, computing aspects, interfacing, and storage. A weighting factor was applied to each criterion of the grid to consider its relative importance. To assess the robustness of the software, six representative clinical vignettes were also processed through all of them.¦Results: 12 software tools were identified, tested and ranked. It represents a comprehensive review of the available software characteristics. Numbers of drugs handled vary from 2 to more than 180, and integration of different population types is available for some programs. Nevertheless, 8 programs offer the ability to add new drug models based on population PK data. 10 computer tools incorporate Bayesian computation to predict dosage regimen (individual parameters are calculated based on population PK models). All of them are able to compute Bayesian a posteriori dosage adaptation based on a blood concentration while 9 are also able to suggest a priori dosage regimen, only based on individual patient covariates. Among those applying Bayesian analysis, MM-USC*PACK uses a non-parametric approach. The top 2 programs emerging from this benchmark are MwPharm and TCIWorks. Others programs evaluated have also a good potential but are less sophisticated or less user-friendly.¦Conclusions: Whereas 2 software packages are ranked at the top of the list, such complex tools would possibly not fit all institutions, and each program must be regarded with respect to individual needs of hospitals or clinicians. Programs should be easy and fast for routine activities, including for non-experienced users. Although interest in TDM tools is growing and efforts were put into it in the last years, there is still room for improvement, especially in terms of institutional information system interfacing, user-friendliness, capability of data storage and automated report generation.