Artificial Intelligence (AI) describes computer systems able to perform tasks that normally requi... more Artificial Intelligence (AI) describes computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. Examples of AI techniques are machine learning, neural networks, and deep learning. AI can be applied in many different areas, such as econometrics, biometry, e-commerce, and the automotive industry. In recent years, AI has found its way into healthcare as well, helping doctors make better decisions (“clinical decision support”), localizing tumors in magnetic resonance images, reading and analyzing reports written by radiologists and pathologists, and much more. However, AI has one big risk: it can be perceived as a “black box”, limiting trust in its reliability, which is a very big issue in an area in which a decision can mean life or death. As a result, the term Explainable Artificial Intelligence (XAI) has been gaining momentum. XAI tries to ensure that AI algorithms (and t...
Artificial Intelligence (AI) describes computer systems able to perform tasks that normally requi... more Artificial Intelligence (AI) describes computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. Examples of AI techniques are machine learning, neural networks and deep learning. AI can be applied in many different areas, such as econometrics, biometry, e-commerce and the automotive industry. In recent years, AI has found its way into healthcare as well, helping doctors to make better decisions (‘clinical decision support’), localizing tumors in magnetic resonance images, reading and analyzing reports written by radiologists and pathologists, and much more. However, AI has one big risk: it can be perceived as a ‘black box’, limiting trust in its reliability, which is a very big issue in an area in which a decision can mean life or death. As a result, the term Explainable Artificial Intelligence (XAI) has been gaining momentum. XAI tries to ensure that AI algorithms (and ...
Among medical specialties, laboratory medicine is the largest producer of structured data and mus... more Among medical specialties, laboratory medicine is the largest producer of structured data and must play a crucial role for the efficient and safe implementation of big data and artificial intelligence in healthcare. The area of personalized therapies and precision medicine has now arrived, with huge data sets not only used for experimental and research approaches, but also in the “real world”. Analysis of real world data requires development of legal, procedural and technical infrastructure. The integration of all clinical data sets for any given patient is important and necessary in order to develop a patient-centered treatment approach. Data-driven research comes with its own challenges and solutions. The Findability, Accessibility, Interoperability, and Reusability (FAIR) Guiding Principles provide guidelines to make data findable, accessible, interoperable and reusable to the research community. Federated learning, standards and ontologies are useful to improve robustness of art...
Active surveillance (AS) is an accepted option for the initial management of carefully selected m... more Active surveillance (AS) is an accepted option for the initial management of carefully selected men with localized , well-differentiated prostate cancer who are thought to have a low risk of progression 1–4. AS is broadly described as a management option for patients with low-risk prostate cancer, which involves the postponement or avoidance of invasive treatment, with a switch to curative treatment if evidence is obtained that the patient has an increased risk of disease progression or if the patient expresses preference for it. However, semantic heterogeneity exists in the literature and guidelines. For instance, the specific definitions of the terms AS and watchful waiting (WW) are inconsistent in the published literature and can elicit considerable confusion. The terms AS and WW are frequently used interchangeably, but they refer to very different observational approaches. AS involves the avoidance or postponement of immediate therapy combined with careful surveillance; definiti...
Active surveillance (AS) is broadly described as a management option for men with low-risk prosta... more Active surveillance (AS) is broadly described as a management option for men with low-risk prostate cancer, but semantic heterogeneity exists in both the literature and in guidelines. To address this issue, a panel of leading prostate cancer specialists in the field of AS participated in a consensus-forming project using a modified Delphi method to reach international consensus on definitions of terms related to this management option. An iterative three-round sequence of online questionnaires designed to address 61 individual items was completed by each panel member. Consensus was considered to be reached if ≥70% of the experts agreed on a definition. To facilitate a common understanding among all experts involved and resolve potential ambiguities, a face-to-face consensus meeting was held between Delphi survey rounds two and three. Convenience sampling was used to construct the panel of experts. In total, 12 experts from Australia, France, Finland, Italy, the Netherlands, Japan, t...
In recent years, more and more health data are being generated. These data come not only from pro... more In recent years, more and more health data are being generated. These data come not only from professional health systems, but also from wearable devices. All these data combined form ‘big data’ that can be utilized to optimize treatments for each unique patient (‘precision medicine’). To achieve this precision medicine, it is necessary that hospitals, academia and industry work together to bridge the ‘valley of death’ of translational medicine. However, hospitals and academia often have problems with sharing their data, even though the patient is actually the owner of his/her own health data, and the sharing of data is associated with increased citation rate. Academic hospitals usually invest a lot of time in setting up clinical trials and collecting data, and want to be the first ones to publish papers on this data. The idea that society benefits the most if the patient’s data are shared as soon as possible so that other researchers can work with it, has not taken root yet. There ...
The Movember Global Action Plan (GAP) on active surveillance for low risk prostate cancer include... more The Movember Global Action Plan (GAP) on active surveillance for low risk prostate cancer includes the integrated 30 months activity of 19 institutions in 14 countries in the 5 Movember regions (Australasia, Europe, UK, Canada, and USA). The initiative is also open to other eligible centres. Milestones of the project include a global Active Surveillance (AS) database for clinical, biospecimen, imaging and biomarker data (including a virtual biobank), as well as worldwide tailor-made guidelines on AS and a web-based platform on AS. The database needs to be accessible for integrated analysis on all datasets from all participating institutes.
In the past years the Smith-Waterman sequence comparison algorithm has gained popularity due to i... more In the past years the Smith-Waterman sequence comparison algorithm has gained popularity due to improved implementations and rapidly increasing computing power. However, the quality and sensitivity of a database search is not only determined by the algorithm but also by the statistical significance testing for an alignment. The e-value is the most commonly used statistical validation method for sequence database searching. The CluSTr database and the Protein World database have been created using an alternative statistical significance test: a Z-score based on Monte-Carlo statistics. Several papers have described the superiority of the Z-score as compared to the e-value, using simulated data. We were interested if this could be validated when applied to existing, evolutionary related protein sequences. All experiments are performed on the ASTRAL SCOP database. The Smith-Waterman sequence comparison algorithm with both e-value and Z-score statistics is evaluated, using ROC, CVE and A...
Phylogenetic patterns show the presence or absence of certain genes or proteins in a set of speci... more Phylogenetic patterns show the presence or absence of certain genes or proteins in a set of species. They can also be used to determine sets of genes or proteins that occur only in certain evolutionary branches. Phylogenetic patterns analysis has routinely been applied to protein databases such as COG and OrthoMCL, but not upon gene databases. Here we present a tool named PhyloPat which allows the complete Ensembl gene database to be queried using phylogenetic patterns. PhyloPat is an easy-to-use webserver, which can be used to query the orthologies of all complete genomes within the EnsMart database using phylogenetic patterns. This enables the determination of sets of genes that occur only in certain evolutionary branches or even single species. We found in total 446,825 genes and 3,164,088 orthologous relationships within the EnsMart v40 database. We used a single linkage clustering algorithm to create 147,922 phylogenetic lineages, using every one of the orthologies provided by ...
The transfer of functional annotations from model organism proteins to human proteins is one of t... more The transfer of functional annotations from model organism proteins to human proteins is one of the main applications of comparative genomics. Various methods are used to analyze cross-species orthologous relationships according to an operational definition of orthology. Often the definition of orthology is incorrectly interpreted as a prediction of proteins that are functionally equivalent across species, while in fact it only defines the existence of a common ancestor for a gene in different species. However, it has been demonstrated that orthologs often reveal significant functional similarity. Therefore, the quality of the orthology prediction is an important factor in the transfer of functional annotations (and other related information). To identify protein pairs with the highest possible functional similarity, it is important to qualify ortholog identification methods. To measure the similarity in function of proteins from different species we used functional genomics data, s...
Gene-oriented sequence clusters (transcriptional units) have found many applications in genomics ... more Gene-oriented sequence clusters (transcriptional units) have found many applications in genomics research including the construction of transcriptome maps and identification of splice variants. We developed a new method to construct transcriptional that uses the genomic sequence as a template. We present and discuss our method in detail together with an evaluation of the transcriptional units for human. We constructed 33,007 and 27,792 transcriptional units for human and mouse, respectively. The sensitivity (81%) and specificity (90%) of our method compares favorably to other established methods. We evaluated the representation of experimentally validated and predicted intergenic spliced transcripts in humans and show that we correctly represent a large fraction of these cases by single transcriptional units. Our method performs well, but the evaluation of the final set of transcriptional units show that improvements to the algorithm are still possible. However, because the precise number and types of errors are difficult to track, it is not obvious how to significantly improve the algorithm. We believe that ongoing research efforts are necessary to further improve current methods. This should include detailed documentation, comparison, and evaluation of current methods.
Artificial Intelligence (AI) describes computer systems able to perform tasks that normally requi... more Artificial Intelligence (AI) describes computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. Examples of AI techniques are machine learning, neural networks, and deep learning. AI can be applied in many different areas, such as econometrics, biometry, e-commerce, and the automotive industry. In recent years, AI has found its way into healthcare as well, helping doctors make better decisions (“clinical decision support”), localizing tumors in magnetic resonance images, reading and analyzing reports written by radiologists and pathologists, and much more. However, AI has one big risk: it can be perceived as a “black box”, limiting trust in its reliability, which is a very big issue in an area in which a decision can mean life or death. As a result, the term Explainable Artificial Intelligence (XAI) has been gaining momentum. XAI tries to ensure that AI algorithms (and t...
Artificial Intelligence (AI) describes computer systems able to perform tasks that normally requi... more Artificial Intelligence (AI) describes computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. Examples of AI techniques are machine learning, neural networks and deep learning. AI can be applied in many different areas, such as econometrics, biometry, e-commerce and the automotive industry. In recent years, AI has found its way into healthcare as well, helping doctors to make better decisions (‘clinical decision support’), localizing tumors in magnetic resonance images, reading and analyzing reports written by radiologists and pathologists, and much more. However, AI has one big risk: it can be perceived as a ‘black box’, limiting trust in its reliability, which is a very big issue in an area in which a decision can mean life or death. As a result, the term Explainable Artificial Intelligence (XAI) has been gaining momentum. XAI tries to ensure that AI algorithms (and ...
Among medical specialties, laboratory medicine is the largest producer of structured data and mus... more Among medical specialties, laboratory medicine is the largest producer of structured data and must play a crucial role for the efficient and safe implementation of big data and artificial intelligence in healthcare. The area of personalized therapies and precision medicine has now arrived, with huge data sets not only used for experimental and research approaches, but also in the “real world”. Analysis of real world data requires development of legal, procedural and technical infrastructure. The integration of all clinical data sets for any given patient is important and necessary in order to develop a patient-centered treatment approach. Data-driven research comes with its own challenges and solutions. The Findability, Accessibility, Interoperability, and Reusability (FAIR) Guiding Principles provide guidelines to make data findable, accessible, interoperable and reusable to the research community. Federated learning, standards and ontologies are useful to improve robustness of art...
Active surveillance (AS) is an accepted option for the initial management of carefully selected m... more Active surveillance (AS) is an accepted option for the initial management of carefully selected men with localized , well-differentiated prostate cancer who are thought to have a low risk of progression 1–4. AS is broadly described as a management option for patients with low-risk prostate cancer, which involves the postponement or avoidance of invasive treatment, with a switch to curative treatment if evidence is obtained that the patient has an increased risk of disease progression or if the patient expresses preference for it. However, semantic heterogeneity exists in the literature and guidelines. For instance, the specific definitions of the terms AS and watchful waiting (WW) are inconsistent in the published literature and can elicit considerable confusion. The terms AS and WW are frequently used interchangeably, but they refer to very different observational approaches. AS involves the avoidance or postponement of immediate therapy combined with careful surveillance; definiti...
Active surveillance (AS) is broadly described as a management option for men with low-risk prosta... more Active surveillance (AS) is broadly described as a management option for men with low-risk prostate cancer, but semantic heterogeneity exists in both the literature and in guidelines. To address this issue, a panel of leading prostate cancer specialists in the field of AS participated in a consensus-forming project using a modified Delphi method to reach international consensus on definitions of terms related to this management option. An iterative three-round sequence of online questionnaires designed to address 61 individual items was completed by each panel member. Consensus was considered to be reached if ≥70% of the experts agreed on a definition. To facilitate a common understanding among all experts involved and resolve potential ambiguities, a face-to-face consensus meeting was held between Delphi survey rounds two and three. Convenience sampling was used to construct the panel of experts. In total, 12 experts from Australia, France, Finland, Italy, the Netherlands, Japan, t...
In recent years, more and more health data are being generated. These data come not only from pro... more In recent years, more and more health data are being generated. These data come not only from professional health systems, but also from wearable devices. All these data combined form ‘big data’ that can be utilized to optimize treatments for each unique patient (‘precision medicine’). To achieve this precision medicine, it is necessary that hospitals, academia and industry work together to bridge the ‘valley of death’ of translational medicine. However, hospitals and academia often have problems with sharing their data, even though the patient is actually the owner of his/her own health data, and the sharing of data is associated with increased citation rate. Academic hospitals usually invest a lot of time in setting up clinical trials and collecting data, and want to be the first ones to publish papers on this data. The idea that society benefits the most if the patient’s data are shared as soon as possible so that other researchers can work with it, has not taken root yet. There ...
The Movember Global Action Plan (GAP) on active surveillance for low risk prostate cancer include... more The Movember Global Action Plan (GAP) on active surveillance for low risk prostate cancer includes the integrated 30 months activity of 19 institutions in 14 countries in the 5 Movember regions (Australasia, Europe, UK, Canada, and USA). The initiative is also open to other eligible centres. Milestones of the project include a global Active Surveillance (AS) database for clinical, biospecimen, imaging and biomarker data (including a virtual biobank), as well as worldwide tailor-made guidelines on AS and a web-based platform on AS. The database needs to be accessible for integrated analysis on all datasets from all participating institutes.
In the past years the Smith-Waterman sequence comparison algorithm has gained popularity due to i... more In the past years the Smith-Waterman sequence comparison algorithm has gained popularity due to improved implementations and rapidly increasing computing power. However, the quality and sensitivity of a database search is not only determined by the algorithm but also by the statistical significance testing for an alignment. The e-value is the most commonly used statistical validation method for sequence database searching. The CluSTr database and the Protein World database have been created using an alternative statistical significance test: a Z-score based on Monte-Carlo statistics. Several papers have described the superiority of the Z-score as compared to the e-value, using simulated data. We were interested if this could be validated when applied to existing, evolutionary related protein sequences. All experiments are performed on the ASTRAL SCOP database. The Smith-Waterman sequence comparison algorithm with both e-value and Z-score statistics is evaluated, using ROC, CVE and A...
Phylogenetic patterns show the presence or absence of certain genes or proteins in a set of speci... more Phylogenetic patterns show the presence or absence of certain genes or proteins in a set of species. They can also be used to determine sets of genes or proteins that occur only in certain evolutionary branches. Phylogenetic patterns analysis has routinely been applied to protein databases such as COG and OrthoMCL, but not upon gene databases. Here we present a tool named PhyloPat which allows the complete Ensembl gene database to be queried using phylogenetic patterns. PhyloPat is an easy-to-use webserver, which can be used to query the orthologies of all complete genomes within the EnsMart database using phylogenetic patterns. This enables the determination of sets of genes that occur only in certain evolutionary branches or even single species. We found in total 446,825 genes and 3,164,088 orthologous relationships within the EnsMart v40 database. We used a single linkage clustering algorithm to create 147,922 phylogenetic lineages, using every one of the orthologies provided by ...
The transfer of functional annotations from model organism proteins to human proteins is one of t... more The transfer of functional annotations from model organism proteins to human proteins is one of the main applications of comparative genomics. Various methods are used to analyze cross-species orthologous relationships according to an operational definition of orthology. Often the definition of orthology is incorrectly interpreted as a prediction of proteins that are functionally equivalent across species, while in fact it only defines the existence of a common ancestor for a gene in different species. However, it has been demonstrated that orthologs often reveal significant functional similarity. Therefore, the quality of the orthology prediction is an important factor in the transfer of functional annotations (and other related information). To identify protein pairs with the highest possible functional similarity, it is important to qualify ortholog identification methods. To measure the similarity in function of proteins from different species we used functional genomics data, s...
Gene-oriented sequence clusters (transcriptional units) have found many applications in genomics ... more Gene-oriented sequence clusters (transcriptional units) have found many applications in genomics research including the construction of transcriptome maps and identification of splice variants. We developed a new method to construct transcriptional that uses the genomic sequence as a template. We present and discuss our method in detail together with an evaluation of the transcriptional units for human. We constructed 33,007 and 27,792 transcriptional units for human and mouse, respectively. The sensitivity (81%) and specificity (90%) of our method compares favorably to other established methods. We evaluated the representation of experimentally validated and predicted intergenic spliced transcripts in humans and show that we correctly represent a large fraction of these cases by single transcriptional units. Our method performs well, but the evaluation of the final set of transcriptional units show that improvements to the algorithm are still possible. However, because the precise number and types of errors are difficult to track, it is not obvious how to significantly improve the algorithm. We believe that ongoing research efforts are necessary to further improve current methods. This should include detailed documentation, comparison, and evaluation of current methods.
Uploads
Papers by Tim Hulsen