SEARCH · PubMed Health

Results for “Algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 973 records · Page 54Linked to original sources

PubMed2001

How useful is a name-based algorithm in health research among Turkish migrants in Germany?

Migrants often face particular social, economic and health disadvantages relative to the population of the host country. In order to adapt health services to the needs of migrants, health researchers need to identify differences in risk factor and disease profiles, as well as inequalities concerning treatment and prevention. Registries of health-related events could be employed for these purposes. In Germany, however, routine data bases often hold no, or inaccurate, information on the national origin of the cases registered. We developed an algorithm based on a large data set of Turkish family and first names (n=15 000), with religion as additional criterion, to identify cases of Turkish origin in registries in a largely automatic search. We tested the performance of the algorithm in a population registry and in a cancer registry. The algorithm discriminates well against Greek and Arab names, with 1% false positive matches in our study. It achieves a specificity of > 99.9% in delimiting Turkish from German cases in the cancer registry. The sensitivity can be increased to 85%, provided the small proportion of case records with uncertain origin can be assessed manually. The name algorithm can be useful for registry-based health research among Turkish migrants in Germany. Possible applications are e.g. in cancer registries to compare survival among German and Turkish cancer patients, or in health insurance registries to compare the relative importance of work-related degenerative diseases. In specific circumstances, the algorithm may also be useful in aetiological research.

Algorithms↗

PubMed2002

Use of clinical algorithms for diagnosing malaria.

Several attempts have been made to identify symptoms and signs based algorithms for diagnosing malaria. In this paper, we review the results of published studies and assess the risks and benefits of this approach in different epidemiological settings. Although in areas with a low prevalence the risk of failure to treat malaria resulting from the use of algorithms was low, the reduction in the wastage of drugs was trivial. The odds of wastage of drugs increased by 1.49 (95% confidence limit 1.45-1.51) for each 10% decrease in the prevalence of malaria. In highly endemic areas the algorithms had a high risk of failure to treat malaria. The odds of failure to treat increased by 1.57 (95% confidence limit 1.50-1.65) for each 10% increase in the prevalence. Furthermore, the best clinical algorithms for diagnosing malaria were site-specific. We conclude that the accuracy of clinical algorithms for diagnosing malaria is not sufficient to determine whether antimalarial drugs should be given to children presenting with febrile illness. In highly endemic areas where laboratory support is not available, the policy of offering antimalarial drugs to all children presenting with a febrile illness recommended by the integrated child management initiative is appropriate.

Algorithms↗

PubMed2001

Inappropriate detection of supraventricular arrhythmias by implantable dual chamber defibrillators: a comparison of four different algorithms.

Inappropriate therapy of supraventricular tachyarrhythmias by an ICD is still a common problem. Dual chamber (DDD) ICDs provide additional atrial sensing and should result in higher specificity for detection of supraventricular tachyarrhythmias. However, a direct comparison of different dual chamber algorithms has not been reported. The detection algorithms of four different DDD ICDs were tested: Phylax AV, Defender IV, Ventak AV III DR, and Gem DR 7271. Based on arrhythmias recorded from patients undergoing invasive electrophysiological studies and in many cases of catheter ablation at our institution, a library consisting of 71 supraventricular and 15 ventricular tachyarrhythmias was created. The library consists of episodes of atrial fibrillation, atrial flutter with different AV conduction, typical and atypical AV nodal reentrant tachycardia, AV reentrant tachycardia, sinus tachycardia, and ventricular tachycardia with and without ventriculoatrial conduction. Atrial fibrillation was appropriately classified by all four algorithms. However, the specificity for detection of other supraventricular tachyarrhythmias achieved by the Biotronik (12%) and the Guidant (11%) devices was significantly lower compared to the specificity of the ELA (28%) and the Medtronic DDD ICD (20%). This is due to the fact that the Biotronik and the Guidant algorithm classified all supraventricular tachyarrhythmias resulting in a stable ventricular rate as ventricular tachycardia, whereas the ELA and Medtronic algorithms performed a more detailed analysis by assessment of PR association, atrial onset, or timing of the atrial event relative to the ventricular event, respectively. Atrial fibrillation, the most common supraventricular tachyarrhythmia in patients with ICD, was detected by all devices.

Algorithms↗

PubMed2003

Discrepant results in the interpretation of HIV-1 drug-resistance genotypic data among widely used algorithms.

OBJECTIVES: The aim of this study was to assess the concordance on the interpretation of HIV-1 drug-resistance genotypic data by three widely used algorithms: Stanford University Database (SU), TruGene (Visible Genetics, Canada) (VG) and VirtualPhenotype (Virco, Belgium) (VP). METHODS: Genotypic data from 293 HIV-1-infected individuals with treatment failure was interpreted for 14 antiretroviral drugs by the three algorithms. RESULTS: Complete concordant results among the three systems for all the drugs studied were found in 40/293 (13.7%) samples. Low concordance in the interpretation was observed for most nucleoside reverse transcriptase inhibitors (NRTIs), while results agreed highly for all nonnucleoside reverse transcriptase inhibitors (NNRTIs) and most protease inhibitors (PIs). In pair-wise comparisons, discordant interpretations between SU and VP were found in over 50% of the samples for didanosine, zalcitabine, stavudine and abacavir, and the level of disagreement between VG and VP exceeded 40% for the same drugs. Major discrepancies (high-level resistance interpretation by one algorithm with sensitive interpretation by another) were observed between VG and VP in over 10% of the cases for didanosine, zalcitabine, stavudine and abacavir. On the other hand, the three algorithms had concordant results for lamivudine in over 90% of the cases. CONCLUSIONS: This work demonstrates the great level of discordance in the interpretation of genotyping results among algorithms, clearly showing the necessity for clinical validation. Moreover, these results suggest that a joint effort from the scientific community as well as national and international HIV societies is needed to achieve a consensus for the interpretation of genotypic data.

Algorithms↗

PubMed2000

Predicting acute renal failure after coronary bypass surgery: cross-validation of two risk-stratification algorithms.

BACKGROUND: Acute renal failure (ARF) requiring dialysis after coronary artery bypass grafting (CABG) occurs in 1 to 5% of patients and is independently associated with postoperative mortality, even after case-mix adjustment. A risk-stratification algorithm that could reliably identify patients at increased risk of ARF could help improve outcomes. METHODS: To assess the validity and generalizability of a previously published preoperative renal risk-stratification algorithm, we analyzed data from the Quality Measurement and Management Initiative (QMMI)1 patient cohort. The QMMI includes all adult patients (N = 9498) who underwent CABG at 1 of 12 academic tertiary care hospitals from August 1993 to October 1995. ARF requiring dialysis was the outcome of interest. Cross-validation of a recursive partitioning algorithm developed from the VA Continuous Improvement in Cardiac Surgery Program (CICSP) was performed on the QMMI. An additive severity score derived from logistic regression was also cross-validated on the QMMI. RESULTS: The CICSP recursive partitioning algorithm discriminated well (ARF vs. no ARF) in QMMI patients, even though the QMMI cohort was more diverse. Rates of ARF were similar among risk subgroups in the CICSP tree, as was the overall ranking of subgroups by risk. Using logistic regression, independent predictors of ARF in the QMMI cohort were similar to those found in the CICSP. The CICSP additive severity score performed well in the QMMI cohort, successfully stratifying patients into low-, medium-, high-, and very high-risk groups. CONCLUSIONS: The CICSP preoperative renal-risk algorithms are valid and generalizable across diverse populations.

Acute Kidney Injury↗

PubMed2002

Physiologic control algorithms for rotary blood pumps using pressure sensor input.

Hierarchical algorithms have been developed for enhanced physiologic control and monitoring of blood pumps using pressure inputs. Pressures were measured at pump inlet and outlet using APEX pressure sensors (APSs). The APS is a patented, long-term implantable, flow-through blood pressure sensor and designed to control implantable heart pumps. The algorithms have been tested using a Donavan circulatory mock-loop setup, a generic rotary pump, and LabVIEW software. The hierarchical algorithms control pump speed using pump inlet pressure as a primary independent variable and pump outlet pressure as a secondary dependent variable. Hierarchical control algorithms based on feedback from pressure sensors can control the speed of the pump to stably maintain ventricular filling pressures and arterial pressures. Monitoring algorithms based on pressure inputs are able to approximate flow rate and hydraulic power for the pump and the left ventricle.

Algorithms↗

PubMed1994

The effect of an intraoperative treatment algorithm on physicians' transfusion practice in cardiac surgery.

BACKGROUND: Inappropriate transfusion in cardiac surgery may, in part, be due to empiric transfusion therapy instituted in the absence of timely laboratory data. Therefore, the effect of a transfusion decision algorithm based on intraoperative coagulation monitoring of physicians' transfusion practice and the transfusion outcome was evaluated. STUDY DESIGN AND METHODS: In a randomized, controlled trial, cardiac surgical patients determined to have microvascular bleeding at the cessation of cardiopulmonary bypass were assigned to algorithm (A) or standard (S) therapy. Group A was treated with plasma and platelet therapy according to a transfusion algorithm based on on-site coagulation data available within 4 minutes. For Group S, the use of laboratory-based data and the decision to transfuse blood components were at physician discretion. RESULTS: Sixty-six patients were entered into the study (Group A, n = 30; Group S, n = 36). Other than the fact that there were significantly more female patients in Group S than in Group A, no differences between cohorts in regard to perioperative risk factors for blood transfusion needs were identified. Therefore, gender was factored in as a covariate in the statistical analysis. Group A patients received fewer hemostatic blood component units (p = 0.008) and had fewer total donor exposures (p = 0.007) during the entire hospitalization period. Linear regression analysis of the differences in slopes in Groups A and S for the relationships between the red cell volume lost and the red cell volume transfused (p < 0.03), non-red cell units transfused (p < 0.0001), and total number of blood components transfused (p < 0.0001) demonstrated that physicians' transfusion practice was significantly altered by the use of a transfusion algorithm with on-site coagulation data, independent of surgical blood losses. CONCLUSION: The use of algorithms by transfusion decision makers can serve as an effective physician education intervention.

Adult↗

PubMed2000

A cell-kinetic model of CD34+ cell mobilization and harvest: development of a predictive algorithm for CD34+ cell yield in PBPC collections.

BACKGROUND: Mobilization and homing of PBPCs are still poorly understood. Thus, a sufficient algorithm for the prediction of PBPC yield in apheresis procedures does not yet exist. STUDY DESIGN AND METHODS: The decline of CD34+ cells in the peripheral blood during apheresis and their simultaneous increase in the collection bag were determined in a prospective study of 18 consecutive apheresis procedures. A cell-kinetic, four-compartment model describing these changes was developed. Retrospective data from 136 apheresis procedures served to further improve this model. A predictive algorithm for the yield was developed that considered the sex, weight, and height of the patient, the number of CD34+ cells in peripheral blood before apheresis, the inlet flow, and the duration of the apheresis. The accuracy of this algorithm was evaluated by comparison of the predicted and the observed yields of CD34+ cells in 105 prospective autologous and 148 retrospective allogeneic apheresis procedures. RESULTS: The correlation between predicted and observed yields was good for the autologous and allogeneic groups with a correlation coefficient (r) of 0.8979 and 0.8311 (p<0.0001), respectively. The regression is described by the equations log (measured value [m]) = 1.0118 + 0.8595 x log (predicted value [p]) for the autologous and log (m) = 2.226 + 0.7559 x log (p) for the allogeneic group. The respective equations for the zero-point regression are log (m) = 1.014 x log (p) and log (m) = 1.026 x log (p). The probability that the measured value was 90 percent or more of the predicted value was 83.8 percent for the autologous and 90.5 percent for the allogeneic apheresis procedures. CONCLUSION: The predictive accuracy of the algorithm and the slope of the zero-point regression curve were higher for allogeneic than autologous PBPC collections. The predictive algorithm may be a useful tool in PBPC harvest, enabling the adaptation of the size of the apheresis to the needs of each patient.

Algorithms↗

PubMed2003

Combined efficacy of atrial septal lead placement and atrial pacing algorithms for prevention of paroxysmal atrial tachyarrhythmia.

INTRODUCTION: The combined role of atrial septal lead location and atrial pacing algorithms in the prevention of atrial tachyarrhythmias (AT/AF), including both atrial fibrillation and flutter, is unknown. We tested the hypothesis that atrial prevention pacing algorithms could decrease AT/AF frequency in patients with atrial septal leads, bradycardia, and paroxysmal AT/AF. METHODS AND RESULTS: A total of 298 patients (age 70 +/- 10 years; 61% male) from 35 centers were implanted with a DDDRP pacing system including three AT/AF prevention pacing algorithms. Lead site was randomized at implant to right atrial septal or nonseptal. Patients were randomized 1 month postimplant to AT/AF prevention ON or OFF for 3 months and then crossed over for 3 months. Patients logged symptomatic AT/AF episodes via a manual activator. Prevention efficacy was evaluated based on intention-to-treat in 277 patients (138 septal) with complete follow-up. No changes in device-recorded AT/AF frequency or burden were observed with algorithms OFF versus ON or between patients randomized to septal versus nonseptal lead location. Analysis of other secondary outcomes revealed that AT/AF prevention pacing resulted in decreased atrial premature contractions in both the septal (1.9 [0.2-8.7] vs 3.3 [0.3-10.6]x 103/day; P < 0.01) and nonseptal groups (0.9 [0.2-3.3] vs 1.3 [0.3-5.5]x 103/day; P < 0.001). Patients with septal leads had fewer symptomatic AT/AF episodes ON versus OFF (1.4 +/- 3.0 vs 2.5 +/- 5.2/month, P = 0.01). CONCLUSION: The combination of three atrial prevention pacing algorithms did not decrease device classified atrial tachyarrhythmia frequency or burden during a 3-month cross-over period in bradycardic patients and septal or nonseptal atrial pacing leads. Prevention pacing was associated with decreased frequency of premature atrial contractions and with decreased symptomatic atrial tachyarrhythmia frequency in patients with atrial septal leads.

Aged↗

PubMed2003

Development and validation of an ECG algorithm for identifying the optimal ablation site for idiopathic ventricular outflow tract tachycardia.

INTRODUCTION: Idiopathic ventricular outflow tract tachycardia or premature ventricular contractions (OT-VTs) can originate from several different sites in the outflow tract, including the left ventricular (LV) endocardium and epicardium. The aims of this study were (1) to develop an ECG algorithm to predict the origin of OT-VT and (2) to test prospectively the accuracy of the algorithm. METHODS AND RESULTS: An algorithm was developed by correlating the 12-lead ECG findings with the catheter ablation site in 80 patients with OT-VT. The ECG characteristics of the QRS complex during the arrhythmia were analyzed. The catheter sites were verified by multi-plane fluoroscopy. The outflow tract was classified into six subdivisions: right ventricular (RV) septum, RV free wall, RV near the His-bundle region, LV endocardium, left sinus of Valsalva (LSV), and LV epicardium remote from the LSV. An OT-VT originating from the LV epicardium remote from the LSV was defined as an OT-VT in which the earliest ventricular activation was recorded at the LSV and radiofrequency ablation from the LSV failed. This algorithm subsequently was tested prospectively in 88 patients. Overall sensitivity was 88% and specificity was 95%. The positive and negative predictive values were 88% and 96%, respectively. CONCLUSION: We describe a new ECG algorithm having a high sensitivity and specificity to identify the optimal ablation site for idiopathic ventricular outflow tachycardia or premature ventricular contractions.

Adolescent↗

PubMed1998

Single-beat analysis of ventricular late potentials in the surface electrocardiogram using the spectrotemporal pattern recognition algorithm in patients with coronary artery disease.

AIMS: Post-infarction risk stratification can be ascertained from beat-to-beat variations in ventricular late potentials. However, gaining such information by conventional late potential analysis using signal averaging is still not possible. METHODS: We therefore developed the spectrotemporal pattern recognition algorithm in order to detect beat-to-beat variations in late potentials. Based on the spectrotemporal pattern recognition algorithm two-dimensional correlation function, the typical spectral pattern of late potentials can be identified in spectrotemporal maps of single beats, even in the presence of noise. RESULTS: Surface electrocardiograms of 385 patients after myocardial infarction (85 with documented sustained ventricular tachycardia (group 1), 100 with fast, polymorphic ventricular tachycardia (> 270 cycles.min-1) or primary ventricular fibrillation (group 2), 200 without ventricular arrhythmias (group 3) and 45 healthy volunteers (group 4), were analysed. The spectrotemporal pattern recognition algorithm detected late potentials in single beats in 89% of group 1 patients, in 79% of group 2, in 22% of group 3 and in 4% of normals. The spectrotemporal pattern recognition algorithm measured late potential frequency and extension of late potentials into the ST segment, which was significantly different between groups 1 and 2. Beat-to-beat variations in late potentials, with respect to frequency and extension into the ST segment, were markedly higher in patients with a history of primary ventricular fibrillation. CONCLUSION: Single-beat analysis using the spectrotemporal pattern recognition algorithm may improve risk stratification of patients after myocardial infarction, and provides information on patients prone to ventricular fibrillation.

Action Potentials↗

PubMed2002

A software-based pacemaker pulse detection and paced rhythm classification algorithm.

A new pacemaker pulse detection and paced electrocardiogram (ECG) rhythm classification algorithm with high sensitivity and positive predictive value has been implemented as part of the Philips Medical Systems' (Andover, MA) ECG analysis program. The detection algorithm was developed on 1,108 paced ECGs with 16,029 individual pulse locations. It operates on 12-lead, 500 sample per second, 150 Hz low-pass filtered ECG signals. Even after low-pass filtering, this algorithm distinguishes between pacemaker pulses and narrow QRS complexes from newborns. An individual pulse detection sensitivity of 99.7% and positive predictive value of 99.5% was obtained by the multi-lead detector. A 10-second, 12-lead ECG database (n = 13,155) of paced (n = 2,190), non-paced adult (n = 8,070), non-paced pediatric (n = 1,209) and "noisy" ECGs with spike noise and muscle artifact (n = 1,686) was assembled and annotated by two readers. The overall performance in identification of an ECG as paced with any pacing present versus non-paced is 97.2% in sensitivity and 99.9% in specificity. The paced ECGs were classified by the mode in which the beats were paced, such as, atrial, ventricular, A-V dual, or dual/inhibited chamber (ie, combinations of atrial, ventricular and dual) pacing. An algorithm was developed for paced rhythm classification. The algorithm performance results show that accurate and robust pacemaker pulse detection and classification can be done in software on diagnostic bandwidth ECG signals.

Adult↗

PubMed2001

Prospective analysis of a fever evaluation algorithm after major gynecologic surgery.

OBJECTIVE: We performed a prospective trial to evaluate the feasibility, accuracy, and safety of a postoperative fever algorithm that is based on symptoms and physical examination in an attempt to decrease the random use of urine cultures, blood cultures, and chest radiographs. STUDY DESIGN: Our fever algorithm consisted of assessing all febrile postoperative patients for signs and symptoms of infection. If none were present, no tests were ordered. RESULTS: Twenty-eight of 105 consecutive patients (27%) had postoperative fever after major gynecologic surgery. Three of 28 febrile patients (11%) were evaluated with tests according to the algorithm. Two of 28 febrile patients (7%) were evaluated in violation of the algorithm. Four febrile patients (14%) had documented infections. Two patients had infections within the first 30 days after discharge. Compared with our previous retrospective review, significantly fewer febrile patients were evaluated with testing with a significantly increased yield of positive test results. CONCLUSIONS: Our postoperative fever evaluation algorithm that is based on symptoms and physical examination is feasible, is safe, decreases random testing, and increases the yield of positive test results.

Adult↗

PubMed2004

A versatile statistical analysis algorithm to detect genome copy number variation.

We have developed a versatile statistical analysis algorithm for the detection of genomic aberrations in human cancer cell lines. The algorithm analyzes genomic data obtained from a variety of array technologies, such as oligonucleotide array, bacterial artificial chromosome array, or array-based comparative genomic hybridization, that operate by hybridizing with genomic material obtained from cancer and normal cells and allow detection of regions of the genome with altered copy number. The number of probes (i.e., resolution), the amount of uncharacterized noise per probe, and the severity of chromosomal aberrations per chromosomal region may vary with the underlying technology, biological sample, and sample preparation. Constrained by these uncertainties, our algorithm aims at robustness by using a priorless maximum a posteriori estimator and at efficiency by a dynamic programming implementation. We illustrate these characteristics of our algorithm by applying it to data obtained from representational oligonucleotide microarray analysis and array-based comparative genomic hybridization technology as well as to synthetic data obtained from an artificial model whose properties can be varied computationally. The algorithm can combine data from multiple sources and thus facilitate the discovery of genes and markers important in cancer, as well as the discovery of loci important in inherited genetic disease.

Algorithms↗

PubMed2002

A dynamic programming algorithm for haplotype block partitioning.

We develop a dynamic programming algorithm for haplotype block partitioning to minimize the number of representative single nucleotide polymorphisms (SNPs) required to account for most of the common haplotypes in each block. Any measure of haplotype quality can be used in the algorithm and of course the measure should depend on the specific application. The dynamic programming algorithm is applied to analyze the chromosome 21 haplotype data of Patil et al. [Patil, N., Berno, A. J., Hinds, D. A., Barrett, W. A., Doshi, J. M., Hacker, C. R., Kautzer, C. R., Lee, D. H., Marjoribanks, C., McDonough, D. P., et al. (2001) Science 294, 1719-1723], who searched for blocks of limited haplotype diversity. Using the same criteria as in Patil et al., we identify a total of 3,582 representative SNPs and 2,575 blocks that are 21.5% and 37.7% smaller, respectively, than those identified using a greedy algorithm of Patil et al. We also apply the dynamic programming algorithm to the same data set based on haplotype diversity. A total of 3,982 representative SNPs and 1,884 blocks are identified to account for 95% of the haplotype diversity in each block.

Algorithms↗

PubMedDate not supplied

Algorithms for network analysis in systems-ADME/Tox using the MetaCore and MetaDrug platforms.

The authors have previously applied two integrated platforms, MetaCore and MetaDrug, for the assembly and analysis of human biological networks as a useful method for the integration and functional interpretation of high-throughput experimental data. The present study demonstrates in detail the specific algorithms that are used in both software platforms. Using a standard set of genes as input, namely CYP3A4 (an enzyme), PXR (a nuclear hormone receptor), MDR1 (a transporter) and hERG (an ion channel) related to the absorption, distribution, metabolism, excretion and toxicity (ADME/Tox) of xenobiotics, we have now generated networks with each algorithm. The relative advantages and disadvantages of these algorithms are explained using these examples as well as appropriate instances of utility to illustrate further the particular circumstances for their use. In addition, the benefits of the different network algorithms are identified when compared with algorithms available in other products, where this information is available.

Algorithms↗

PubMed1991

An algorithm for the DNA sequence generation from k-tuple word contents of the minimal number of random fragments.

An algorithm is described for generation of the long sequence written in a four letter alphabet from the constituent k-tuple words in the minimal number of separate, randomly defined fragments of the starting sequence. It is primarily intended for use in sequencing by hybridization (SBH) process- a potential method for sequencing human genome DNA (Drmanac et al., Genomics 4, pp. 114-128, 1989). The algorithm is based on the formerly defined rules and informative entities of the linear sequence. The algorithm requires neither knowledge on the number of appearances of a given k-tuple in sequence fragments, nor the information on which k-tuple words are on the ends of a fragment. It operates with the mixed content of k-tuples of the various lengths. The concept of the algorithm enables operations with the k-tuple sets containing false positive and false negative k-tuples. The content of the false k-tuples primarily affects the completeness of the generated sequence, and its correctness in the specific cases only. The algorithm can be used for the optimization of SBH parameters in the simulation experiments, as well as for the sequence generation in the real SBH experiments on the genomic DNA.

Algorithms↗

PubMed2005

Evaluating intraspecific "network" construction methods using simulated sequence data: do existing algorithms outperform the global maximum parsimony approach?

In intraspecific studies, reticulated graphs are valuable tools for visualization, within a single figure, of alternative genealogical pathways among haplotypes. As available software packages implementing the global maximum parsimony (MP) approach only give the possibility to merge resulting topologies into less-resolved consensus trees, MP has often been neglected as an alternative approach to purely algorithmic (i.e., methods defined solely on the basis of an algorithm) "network" construction methods. Here, we propose to search tree space using the MP criterion and present a new algorithm for uniting all equally most parsimonious trees into a single (possibly reticulated) graph. Using simulated sequence data, we compare our method with three purely algorithmic and widely used graph construction approaches (minimum-spanning network, statistical parsimony, and median-joining network). We demonstrate that the combination of MP trees into a single graph provides a good estimate of the true genealogy. Moreover, our analyses indicate that, when internal node haplotypes are not sampled, the median-joining and MP methods provide the best estimate of the true genealogy whereas the minimum-spanning algorithm shows very poor performances.

Algorithms↗