Google Bard and ChatGPT in Orthopedics: Which Is the Better Doctor in Sports Medicine and Pediatric Orthopedics? The Role of AI in Patient Education.

BACKGROUND: This study evaluates the potential of ChatGPT and Google Bard as educational tools for patients in orthopedics, focusing on sports medicine and pediatric orthopedics. The aim is to compare the quality of responses provided by these natural language processing (NLP) models, addressing concerns about the potential dissemination of incorrect medical information. METHODS: Ten ACL- and flat foot-related questions from a Google search were presented to ChatGPT-3.5 and Google Bard. Expert orthopedic surgeons rated the responses using the Global Quality Score (GQS). The study minimized bias by clearing chat history before each question, maintaining respondent anonymity and employing statistical analysis to compare response quality. RESULTS: ChatGPT-3.5 and Google Bard yielded good-quality responses, with average scores of 4.1 ± 0.7 and 4 ± 0.78, respectively, for sports medicine. For pediatric orthopedics, Google Bard scored 3.5 ± 1, while the average score for responses generated by ChatGPT was 3.8 ± 0.83. In both cases, no statistically significant difference was found between the platforms (p = 0.6787, p = 0.3092). Despite ChatGPT's responses being considered more readable, both platforms showed promise for AI-driven patient education, with no reported misinformation. CONCLUSIONS: ChatGPT and Google Bard demonstrate significant potential as supplementary patient education resources in orthopedics. However, improvements are needed for increased reliability. The study underscores the evolving role of AI in orthopedics and calls for continued research to ensure a conscientious integration of AI in healthcare education.

Limb asymmetries persist 6 months after anterior cruciate ligament reconstruction according to the results of a jump test battery.

Objectives: Test batteries used to assess a patient's return-to-sports (RTS) following anterior cruciate ligament reconstruction (ACLR) are currently undergoing continual development, although no consensus exist on tests to be administered to athletes before allowing return to play. A simple standardized jump test battery was developed to objectively evaluate knee function following ACLR, thereby aiding in RTS decision-making. Methods: Thirty-three patients who underwent ACLR were prospectively assessed pre-operatively, 6, and 12 months after surgery. Knee function was assessed using a device for optical detection using a test battery consisting of three jump tests: monopodalic countermovement jump (CMJ), drop jump, and monopodalic side-hop. Limb symmetry index (LSI) was reported for all tests at all time points. LSI ≥90% was defined as RTS criteria. Results: At 12-month evaluation, mean LSI significantly improved compared to 6-month follow up (p < 0.01), and also compared to baseline (p < 0.01), reporting a mean value of 92.6% for CMJ, 90.6 for drop jump and 96.9% for side hop test. Most patients fulfilled the RTS criteria 12 months after surgery (LSI ≥90%). The percentages of patients demonstrating LSI ≥90% at 6 months was 7/33 (21.2%) for CMJ, 12/33 (36.4%) for drop jump, and 11/33 (33.3%) for side-hop test. One year after surgery, percentages grew up to 66.6% (22/33), 63.6% (21/33), and 81.8% (27/33) respectively. Conclusion: Six months after ACLR, knee functional performance was unsatisfactory in most patients, whereas a significantly higher percentage of patients met RTS criteria 1 year after surgery. The results of the jump test battery proposed in this study support the idea that timing for resumption of cutting and pivoting sports should be delayed later than 6 months, as still limb asymmetries persist at this time point.

Chronotype and quality of sleep in alpine skiers.

Alpine skiing is among the most demanding sporting activities in terms of physical effort and mental workload. The aim of the study was to compare sleep quality and chronotype distribution between 84 highly trained alpine skiers and a control sample of 84 non-athletes matched by age and sex ratio. Quality of sleep was assessed by the Pittsburgh Quality of Sleep Index (PSQI), and chronotype was assessed by the Morningness-Eveningness Questionnaire (MEQ). Additional questions assessed sleep management during training or competitions. The results showed a marked skewed chronotype distribution towards morningness in alpine skiers (52.4% morning type, 42.8% intermediate, and 4.8% evening type) in comparison to the control group. The midpoint of sleep was significantly anticipated among alpine skiers. Differently from the previous literature that showed poor sleep quality and quantity in competitive athletes, the quality and quantity of sleep in alpine skiers was within the normal range in all the PSQI subcomponents.

Drop Jump Performance Improves One Year Following Anterior Cruciate Ligament Reconstruction in Sportsmen Irrespectively of Psychological Patient Reported Outcomes.

Our study aims to prospectively report the functional outcomes of 31 sportsmen following anterior cruciate ligament (ACL) reconstruction, up to 12 months after surgery, with regards to subjective tests and drop jump performance, and to investigate the correlations between these variables, to be used for determining the return to sports after ACL reconstruction. Lysholm score, Tegner activity level, and the ACL-Return to Sport after Injury (ACL-RSI) scale were evaluated preoperatively, at 6 months, and at 12 months after surgery. Drop vertical jump was recorded using an infrared optical acquisition system. Lysholm and ACL-RSI scores significantly improved at the 12-month follow-up compared to the baseline and 6-month evaluations (p < 0.001). Concerning Tegner activity level, no statistically significant differences were reported between pre- and post-operative status (p = 0.179). Drop jump limb symmetry index significantly improved at 12 months, with the mean value improving from 76.6% (SD: 32,4) pre-operatively to 90.2% (SD: 14.7; p < 0.001) at follow-up. Scarce positive correlation was reported between the ability to perform drop jumps and activity level in athletes one year after ACL reconstruction. In addition, subjective knee score and psychological readiness were not related to jumping performance.

Relationships between Jumping Performance and Psychological Readiness to Return to Sport 6 Months Following Anterior Cruciate Ligament Reconstruction: A Cross-Sectional Study.

Background: Investigating the relationship between functional capacity and psychological readiness is of paramount importance when planning sport resumption following knee surgery. The aim of this study was to prospectively assess clinical and functional outcomes in athletes 6 months after primary anterior cruciate ligament (ACL) reconstruction and to evaluate whether jumping ability is related to psychological readiness to return to sport following ACL surgery. Methods: Patients who underwent ACL reconstruction were prospectively enrolled and evaluated pre-operatively and 6 months after surgery. Assessment included Lysholm score, International Knee Documentation Committee (IKDC) Subjective Knee Form, Tegner activity level, and the ACL−Return to Sport after Injury (ACL-RSI) scale. Jumping ability was instrumentally assessed by an infrared optical acquisition system using a test battery including mono- and bipodalic vertical jump and a side hop test. Patients were dichotomized by ACL-RSI into two groups: group A (ACL-RSI > 60), and group B (ACL-RSI < 60). Results: Overall, 29 males and two females from the original study group of 37 patients (84%) were available for clinical evaluation. Mean age at surgery was 34.2 years (SD 11.3). Mean body mass index (BMI) was 25.4 (SD 3.7). Mean overall Lysholm, IKDC, and ACL-RSI scores increased from pre-operatively (p < 0.001). No differences in Tegner score were reported (p = 0.161). Similarly, improvement in most variables regarding jumping ability were observed at follow-up (p < 0.05). According to ACL-RSI, 20 subjects were allocated in group A (ACL-RSI > 60), while 11 were allocated in group B (ACL-RSI < 60). A statistically significant difference in favor of patients in group A was recorded for the post-operative Lysholm and Tegner score, as well as Side Hop test LSI level (p < 0.05), while a trend for IKDC was observed without statistical significance (p = 0.065). Conclusions: Patients with higher values of ACL-RSI scores showed better functional and clinical outcomes as well as improved performance 6 months after ACL reconstruction

First metatarsal extracapsular osteotomy to treat moderate hallux valgus deformity: the modified Wilson-SERI techinique.

From February 2017 to December 2018, 20 patients had undergone the proposed modified Wilson-SERI osteotomy technique, for moderate hallux valgus. The mean age of patients was 58,25 years (range 19 to 78). The hallux valgus angle (HVA), the intermetatarsal angle between first and second metatarsal bone (IMA) and the distal metatarsal articular angle (D.M.A.A) were measured. The feet were assessed based on the scoring system used by Broughton and Winson and by the American Orthopedic Foot and Ankle Society (AOFAS) hallux-metatarsophalangeal-interphalangeal scale. All twenty one patients were followed up postoperatively for a minimum of 12 months. The mean HVA angle decreased significantly from 31,1° before surgery (range 22.9°-40°SD 5.0) at 11,2° (range 2.5° to 22.0°SD 5.3) at twelve months follow up. The mean IMA angle decreased significantly from 12,5° (range 8.0°-18.6°SD 3.8) before surgery at 7,4° (range 3.4°-14.0°SD 2.5) at twelve months follow up. The mean DMMA angle decreased significantly from 15.1° (range 5.3° to 20.0°SD 4.4) before surgery at 7,4 °(1.5°- 10.7°SD 2.5) at twelve months follow up. The mean score according to the AOFAS forefoot was increased from 22,1 (range 13-30 SD 5.0) to 88,2 (Range 77-96 SD 5.2) (p<0.0001). No complications, like dislocations, avascular necrosis of the first metatarsal and deep venous thrombosis, were observed in the post-operative period. Short term results at twelve months after surgery are quite satisfactory but further studies are necessary, to better comprehend an overall outcome of such approach in the long run.

parSMURF, a high-performance computing tool for the genome-wide detection of pathogenic variants.

BACKGROUND: Several prediction problems in computational biology and genomic medicine are characterized by both big data as well as a high imbalance between examples to be learned, whereby positive examples can represent a tiny minority with respect to negative examples. For instance, deleterious or pathogenic variants are overwhelmed by the sea of neutral variants in the non-coding regions of the genome: thus, the prediction of deleterious variants is a challenging, highly imbalanced classification problem, and classical prediction tools fail to detect the rare pathogenic examples among the huge amount of neutral variants or undergo severe restrictions in managing big genomic data. RESULTS: To overcome these limitations we propose parSMURF, a method that adopts a hyper-ensemble approach and oversampling and undersampling techniques to deal with imbalanced data, and parallel computational techniques to both manage big genomic data and substantially speed up the computation. The synergy between Bayesian optimization techniques and the parallel nature of parSMURF enables efficient and user-friendly automatic tuning of the hyper-parameters of the algorithm, and allows specific learning problems in genomic medicine to be easily fit. Moreover, by using MPI parallel and machine learning ensemble techniques, parSMURF can manage big data by partitioning them across the nodes of a high-performance computing cluster. Results with synthetic data and with single-nucleotide variants associated with Mendelian diseases and with genome-wide association study hits in the non-coding regions of the human genome, involhing millions of examples, show that parSMURF achieves state-of-the-art results and an 80-fold speed-up with respect to the sequential version. CONCLUSIONS: parSMURF is a parallel machine learning tool that can be trained to learn different genomic problems, and its multiple levels of parallelization and high scalability allow us to efficiently fit problems characterized by big and imbalanced genomic data. The C++ OpenMP multi-core version tailored to a single workstation and the C++ MPI/OpenMP hybrid multi-core and multi-node parSMURF version tailored to a High Performance Computing cluster are both available at https://github.com/AnacletoLAB/parSMURF.

Network modeling of patients' biomolecular profiles for clinical phenotype/outcome prediction.

Methods for phenotype and outcome prediction are largely based on inductive supervised models that use selected biomarkers to make predictions, without explicitly considering the functional relationships between individuals. We introduce a novel network-based approach named Patient-Net (P-Net) in which biomolecular profiles of patients are modeled in a graph-structured space that represents gene expression relationships between patients. Then a kernel-based semi-supervised transductive algorithm is applied to the graph to explore the overall topology of the graph and to predict the phenotype/clinical outcome of patients. Experimental tests involving several publicly available datasets of patients afflicted with pancreatic, breast, colon and colorectal cancer show that our proposed method is competitive with state-of-the-art supervised and semi-supervised predictive systems. Importantly, P-Net also provides interpretable models that can be easily visualized to gain clues about the relationships between patients, and to formulate hypotheses about their stratification.

Return to sports and re-rupture rate following anterior cruciate ligament reconstruction in amateur sportsman: long-term outcomes.

BACKGROUND: The aim of this study was to retrospectively evaluate patient satisfaction, the return-to-sport rate and activity level at a long-term follow-up in a large cohort of amateur sportsmen who underwent primary anterior cruciate ligament (ACL) reconstruction. METHODS: A total of 218 patients who underwent primary ACL reconstruction between 2004 and 2011, were successfully recontacted and retrospectively reviewed at an average follow-up of 10.5 years (range, 7 to 14 years). All surgeries were performed by one single surgeon. All of them underwent primary ACL reconstruction with autogenous hamstring tendon grafts. Assessment included Knee Osteoarthritis Outcome Score (KOOS) score, International Knee Documentation Committee (IKDC) Subjective Knee Form, Tegner activity level. Patients were also asked what kind of injury they sustained (either direct or indirect trauma), what kind of sport they were performing when they got injured, at what time they did return to sports and which sport they practised before and after surgery. RESULTS: Fourteen patients underwent re-rupture. In 11 cases, this was due to a new trauma occurring at an average time of 22.9 (SD 23.8) months following primary surgery. In 3 cases rupture occurred during rehabilitation period. Mean postoperative KOOS score was 88.5 (SD 8.5), while mean IKDC subjective score was 87.5 (SD 10.9). At the time of follow-up, most patients (214 subjects, 98%) were participating in sport. 156 subjects returned to pre-injury level (71.6%). CONCLUSIONS: The study reported long-term favourable subjective outcomes in amateur sportsman following ACL reconstruction, with a low re-rupture rate and a high percentage of subjects (93.6%) returning to sports participation 12 months after surgery. Most patients (71.6%) were able to return to their preprimary level of activity and sport. Younger age at the time of ACL reconstruction positively affected return to sports; however, younger patients were significantly more likely than older patients to undergo re-rupture.

Minimally Invasive Plate Osteosynthesis for Proximal Humerus Fractures: A Retrospective Study Describing Principles and Advantages of the Technique.

BACKGROUND: The aim of this study was to evaluate the clinical and radiographic results after minimally invasive plate osteosynthesis (MIPO) for proximal humerus fractures. Potential advantages of this approach include the easier exposure of the greater tuberosity and the limited surgical dissection around the fracture site. MATERIALS AND METHODS: From October 2011 to March 2016, thirty-nine patients (32 women, 7 men) with a mean age of 64.9 years (range: 48-80) were surgically treated with the MIPO technique for proximal humeral fractures. According to Neer classification, there were 12 two-part, 24 three-part, and 2 four-part fractures and 1 two-part fracture-dislocation; the AO/OTA system was also used to categorize the fractures. The Constant-Murley (CMS) and the Oxford Shoulder (OSS) Scores were used to evaluate shoulder function. RESULTS: Thirty-four patients were available for clinical and radiographic evaluation at a mean follow-up of 31.8 months (range: 12-54 months). All fractures healed and no postoperative complications occurred. Full recovery of pretrauma activities was reported by 27 patients, while 7 patients presented mild functional limitations. The mean absolute CMS was 75.2 (range: 55-95), the mean normalized CMS was 90.5 (range: 69-107), and the mean OSS was 43.7 (range: 31-48). The only statistically significant correlation was found between the female gender and lower absolute CMS and OSS. Radiographic evaluation revealed varus malunion in 4 cases and valgus malunion in 1 case, while incomplete greater tuberosity reduction was detected in 4 cases. All malunions were related to inadequate reduction at time of surgery and not to secondary displacement. CONCLUSIONS: MIPO for proximal humeral fractures is an effective and safe surgical procedure. The limited tissue dissection allows minimizing the incidence of nonunion, avascular necrosis, and infection. The technique is not easy, requires experience to achieve mastery, and should be reserved for selected fracture patterns. In our experience, the main advantage of this approach consists in the direct access to the greater tuberosity, thus facilitating its anatomic reduction and fixation.

Imbalance-Aware Machine Learning for Predicting Rare and Common Disease-Associated Non-Coding Variants.

Disease and trait-associated variants represent a tiny minority of all known genetic variation, and therefore there is necessarily an imbalance between the small set of available disease-associated and the much larger set of non-deleterious genomic variation, especially in non-coding regulatory regions of human genome. Machine Learning (ML) methods for predicting disease-associated non-coding variants are faced with a chicken and egg problem - such variants cannot be easily found without ML, but ML cannot begin to be effective until a sufficient number of instances have been found. Most of state-of-the-art ML-based methods do not adopt specific imbalance-aware learning techniques to deal with imbalanced data that naturally arise in several genome-wide variant scoring problems, thus resulting in a significant reduction of sensitivity and precision. We present a novel method that adopts imbalance-aware learning strategies based on resampling techniques and a hyper-ensemble approach that outperforms state-of-the-art methods in two different contexts: the prediction of non-coding variants associated with Mendelian and with complex diseases. We show that imbalance-aware ML is a key issue for the design of robust and accurate prediction algorithms and we provide a method and an easy-to-use software tool that can be effectively applied to this challenging prediction task.

RANKS: a flexible tool for node label ranking and classification in biological networks.

UNLABELLED: RANKS is a flexible software package that can be easily applied to any bioinformatics task formalizable as ranking of nodes with respect to a property given as a label, such as automated protein function prediction, gene disease prioritization and drug repositioning. To this end RANKS provides an efficient and easy-to-use implementation of kernelized score functions, a semi-supervised algorithmic scheme embedding both local and global learning strategies for the analysis of biomolecular networks. To facilitate comparative assessment, baseline network-based methods, e.g. label propagation and random walk algorithms, have also been implemented. AVAILABILITY AND IMPLEMENTATION: The package is available from CRAN: https://cran.r-project.org/ The package is written in R, except for the most computationally intensive functionalities which are implemented in C. CONTACT: valentini@di.unimi.it SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Think globally and solve locally: secondary memory-based network learning for automated multi-species function prediction.

BACKGROUND: Network-based learning algorithms for automated function prediction (AFP) are negatively affected by the limited coverage of experimental data and limited a priori known functional annotations. As a consequence their application to model organisms is often restricted to well characterized biological processes and pathways, and their effectiveness with poorly annotated species is relatively limited. A possible solution to this problem might consist in the construction of big networks including multiple species, but this in turn poses challenging computational problems, due to the scalability limitations of existing algorithms and the main memory requirements induced by the construction of big networks. Distributed computation or the usage of big computers could in principle respond to these issues, but raises further algorithmic problems and require resources not satisfiable with simple off-the-shelf computers. RESULTS: We propose a novel framework for scalable network-based learning of multi-species protein functions based on both a local implementation of existing algorithms and the adoption of innovative technologies: we solve "locally" the AFP problem, by designing "vertex-centric" implementations of network-based algorithms, but we do not give up thinking "globally" by exploiting the overall topology of the network. This is made possible by the adoption of secondary memory-based technologies that allow the efficient use of the large memory available on disks, thus overcoming the main memory limitations of modern off-the-shelf computers. This approach has been applied to the analysis of a large multi-species network including more than 300 species of bacteria and to a network with more than 200,000 proteins belonging to 13 Eukaryotic species. To our knowledge this is the first work where secondary-memory based network analysis has been applied to multi-species function prediction using biological networks with hundreds of thousands of proteins. CONCLUSIONS: The combination of these algorithmic and technological approaches makes feasible the analysis of large multi-species networks using ordinary computers with limited speed and primary memory, and in perspective could enable the analysis of huge networks (e.g. the whole proteomes available in SwissProt), using well-equipped stand-alone machines.

An extensive analysis of disease-gene associations using network integration and fast kernel-based gene prioritization methods.

OBJECTIVE: In the context of "network medicine", gene prioritization methods represent one of the main tools to discover candidate disease genes by exploiting the large amount of data covering different types of functional relationships between genes. Several works proposed to integrate multiple sources of data to improve disease gene prioritization, but to our knowledge no systematic studies focused on the quantitative evaluation of the impact of network integration on gene prioritization. In this paper, we aim at providing an extensive analysis of gene-disease associations not limited to genetic disorders, and a systematic comparison of different network integration methods for gene prioritization. MATERIALS AND METHODS: We collected nine different functional networks representing different functional relationships between genes, and we combined them through both unweighted and weighted network integration methods. We then prioritized genes with respect to each of the considered 708 medical subject headings (MeSH) diseases by applying classical guilt-by-association, random walk and random walk with restart algorithms, and the recently proposed kernelized score functions. RESULTS: The results obtained with classical random walk algorithms and the best single network achieved an average area under the curve (AUC) across the 708 MeSH diseases of about 0.82, while kernelized score functions and network integration boosted the average AUC to about 0.89. Weighted integration, by exploiting the different "informativeness" embedded in different functional networks, outperforms unweighted integration at 0.01 significance level, according to the Wilcoxon signed rank sum test. For each MeSH disease we provide the top-ranked unannotated candidate genes, available for further bio-medical investigation. CONCLUSIONS: Network integration is necessary to boost the performances of gene prioritization methods. Moreover the methods based on kernelized score functions can further enhance disease gene ranking results, by adopting both local and global learning strategies, able to exploit the overall topology of the network.

A neural network algorithm for semi-supervised node label learning from unbalanced data.

Given a weighted graph and a partial node labeling, the graph classification problem consists in predicting the labels of all the nodes. In several application domains, from gene to social network analysis, the labeling is unbalanced: for instance positive labels may be much less than negatives. In this paper we present COSNet (COst Sensitive neural Network), a neural algorithm for predicting node labels in graphs with unbalanced labels. COSNet is based on a 2-parameter family of Hopfield networks, and consists of two main steps: (1) the network parameters are learned through a cost-sensitive optimization procedure; (2) a suitable Hopfield network restricted to the unlabeled nodes is considered and simulated. The reached equilibrium point induces the classification of the unlabeled nodes. The restriction of the dynamics leads to a significant reduction in time complexity and allows the algorithm to nicely scale with large networks. An experimental analysis on real-world unbalanced data, in the context of the genome-wide prediction of gene functions, shows the effectiveness of the proposed approach.

Network-based drug ranking and repositioning with respect to DrugBank therapeutic categories.

Drug repositioning is a challenging computational problem involving the integration of heterogeneous sources of biomolecular data and the design of label ranking algorithms able to exploit the overall topology of the underlying pharmacological network. In this context, we propose a novel semisupervised drug ranking problem: prioritizing drugs in integrated biochemical networks according to specific DrugBank therapeutic categories. Algorithms for drug repositioning usually perform the inference step into an inhomogeneous similarity space induced by the relationships existing between drugs and a second type of entity (e.g., disease, target, ligand set), thus making unfeasible a drug ranking within a homogeneous pharmacological space. To deal with this problem, we designed a general framework based on bipartite network projections by which homogeneous pharmacological networks can be constructed and integrated from heterogeneous and complementary sources of chemical, biomolecular and clinical information. Moreover, we present a novel algorithmic scheme based on kernelized score functions that adopts both local and global learning strategies to effectively rank drugs in the integrated pharmacological space using different network combination methods. Detailed experiments with more than 80 DrugBank therapeutic categories involving about 1,300 FDA-approved drugs show the effectiveness of the proposed approach.

A fast ranking algorithm for predicting gene functions in biomolecular networks.

Ranking genes in functional networks according to a specific biological function is a challenging task raising relevant performance and computational complexity problems. To cope with both these problems we developed a transductive gene ranking method based on kernelized score functions able to fully exploit the topology and the graph structure of biomolecular networks and to capture significant functional relationships between genes. We run the method on a network constructed by integrating multiple biomolecular data sources in the yeast model organism, achieving significantly better results than the compared state-of-the-art network-based algorithms for gene function prediction, and with relevant savings in computational time. The proposed approach is general and fast enough to be in perspective applied to other relevant node ranking problems in large and complex biological networks.

Cancer module genes ranking using kernelized score functions.

BACKGROUND: Co-expression based Cancer Modules (CMs) are sets of genes that act in concert to carry out specific functions in different cancer types, and are constructed by exploiting gene expression profiles related to specific clinical conditions or expression signatures associated to specific processes altered in cancer. Unfortunately, genes involved in cancer are not always detectable using only expression signatures or co-expressed sets of genes, and in principle other types of functional interactions should be exploited to obtain a comprehensive picture of the molecular mechanisms underlying the onset and progression of cancer. RESULTS: We propose a novel semi-supervised method to rank genes with respect to CMs using networks constructed from different sources of functional information, not limited to gene expression data. It exploits on the one hand local learning strategies through score functions that extend the guilt-by-association approach, and on the other hand global learning strategies through graph kernels embedded in the score functions, able to take into account the overall topology of the network. The proposed kernelized score functions compare favorably with other state-of-the-art semi-supervised machine learning methods for gene ranking in biological networks and scales well with the number of genes, thus allowing fast processing of very large gene networks. CONCLUSIONS: The modular nature of kernelized score functions provides an algorithmic scheme from which different gene ranking algorithms can be derived, and the results show that using integrated functional networks we can successfully predict CMs defined mainly through expression signatures obtained from gene expression data profiling. A preliminary analysis of top ranked "false positive" genes shows that our approach could be in perspective applied to discover novel genes involved in the onset and progression of tumors related to specific CMs.

Regeneration-associated WNT signaling is activated in long-term reconstituting AC133bright acute myeloid leukemia cells.

Acute myeloid leukemia (AML) is a genetically heterogeneous clonal disorder characterized by two molecularly distinct self-renewing leukemic stem cell (LSC) populations most closely related to normal progenitors and organized as a hierarchy. A requirement for WNT/ß-catenin signaling in the pathogenesis of AML has recently been suggested by a mouse model. However, its relationship to a specific molecular function promoting retention of self-renewing leukemia-initiating cells (LICs) in human remains elusive. To identify transcriptional programs involved in the maintenance of a self-renewing state in LICs, we performed the expression profiling in normal (n = 10) and leukemic (n = 33) human long-term reconstituting AC133(+) cells, which represent an expanded cell population in most AML patients. This study reveals the ligand-dependent WNT pathway activation in AC133(bright) AML cells and shows a diffuse expression and release of WNT10B, a hematopoietic stem cell regenerative-associated molecule. The establishment of a primary AC133(+) AML cell culture (A46) demonstrated that leukemia cells synthesize and secrete WNT ligands, increasing the levels of dephosphorylated ß-catenin in vivo. We tested the LSC functional activity in AC133(+) cells and found significant levels of engraftment upon transplantation of A46 cells into irradiated Rag2(-/-)γc(-/-) mice. Owing to the link between hematopoietic regeneration and developmental signaling, we transplanted A46 cells into developing zebrafish. This system revealed the formation of ectopic structures by activating dorsal organizer markers that act downstream of the WNT pathway. In conclusion, our findings suggest that AC133(bright) LSCs are promoted by misappropriating homeostatic WNT programs that control hematopoietic regeneration.

Noise tolerance of multiple classifier systems in data integration-based gene function prediction.

The availability of various high-throughput experimental and computational methods developed in the last decade allowed molecular biologists to investigate the functions of genes at system level opening unprecedented research opportunities. Despite the automated prediction of genes functions could be included in the most difficult problems in bioinformatics, several recently published works showed that consistent improvements in prediction performances can be obtained by integrating heterogeneous data sources. Nevertheless, very few works have been dedicated to the investigation of the impact of noisy data on the prediction performances achievable by using data integration approaches. In this contribution we investigated the tolerance of multiple classifier systems (MCS) to noisy data in gene function prediction experiments based on data integration methods. The experimental results show that performances of MCS do not undergo a significant decay when noisy data sets are added. In addition, we show that in this task MCS are competitive with kernel fusion, one of the most widely applied technique for data integration in gene function prediction problems.