Tag Archive for Schistosoma

R for Parasitology

How to use R in Parasitology – Schistosoma diagnostics

Recently I started using R-computing. The applications of R are just fantastic. In the future I want to share more of my R-code and its applications in the field of parasitology.

In this example I used some data of some public funded study from a decade ago. Nonetheless, I put noise into the dataset using functions like ‘x <- sample(0:5, 176, replace=T)’ and ‘a <- rnorm(176, sd = 0.6)’. The data is now significantly different than the original data set. All similarities are purely coincidental. I just want to show the method, not the data.

Suppose you have the diagnostic data of microscopic examinations of feces and urine with the number of Schistoma eggs counted. In feces it’s Schistoma mansoni eggs and in urine it’s Schistosoma haematobium. You want to compare these values with a the real-time PCR analyses on the same stool samples. S. haemtobium PCR will also be performed on stool samples (yes, it’s possible).

Make sure the workingspace is correct and import the data from the file: schistoPCR2.txt.


## import dataset in txt file
worms <- read.table("~/Documents/workspace/R-project/SCHISTO/schistoPCR2.txt", 
    header = T, row.names = 1)

# to see variables by name
##      Ct_Sm          Ct_Sh         Ct_PhHV          Kato     
##  Min.   : 0.0   Min.   : 0.0   Min.   :30.4   Min.   :   0  
##  1st Qu.: 0.0   1st Qu.: 0.0   1st Qu.:31.9   1st Qu.:   0  
##  Median :24.2   Median : 0.0   Median :32.7   Median :  80  
##  Mean   :17.9   Mean   :15.0   Mean   :32.7   Mean   : 454  
##  3rd Qu.:27.8   3rd Qu.:35.5   3rd Qu.:33.3   3rd Qu.: 340  
##  Max.   :38.4   Max.   :45.0   Max.   :37.5   Max.   :9320  
##      urine       
##  Min.   :   0.0  
##  1st Qu.:   3.0  
##  Median :   6.0  
##  Mean   :  67.1  
##  3rd Qu.:  40.8  
##  Max.   :1097.9
##   Ct_Sm Ct_Sh Ct_PhHV Kato urine
## 1 38.39     0   34.60  100   4.0
## 2 23.94     0   33.07  340   4.0
## 3 25.31     0   33.45   40   0.0
## 4 31.54     0   34.67    0   0.0
## 5 29.87     0   32.94    0  14.5
## 6 24.35     0   33.64 2160   3.0

Now lets see what the plot looks like when you compare Ct values with Microscopic data.

## correlate Kato & CtSm 
cor(worms$Ct_Sm, worms$Kato)
## [1] 0.1637
plot(worms$Ct_Sm, worms$Kato)

plot of chunk unnamed-chunk-2

It looks awfull. First problem: ‘not-detected’ Ct-values have value zero. It needs to be changed into a ‘weakest’ or ‘highest’ Ct-value in order to keep the negatives in line with the rest of the Ct-values.

Look for the maximum Ct value of CtSm and CtSh. Then replace ‘0’-Ct-values with max value plus 3.3 Ct’s

x <- max(worms[, 1:2], na.rm = TRUE)
## [1] 45

CtM <- x + 3.3
## [1] 48.3

worms$Ct_Sm <- replace(worms$Ct_Sm, worms$Ct_Sm == 0, CtM)
worms$Ct_Sh <- replace(worms$Ct_Sh, worms$Ct_Sh == 0, CtM)

Do the same with the negative microscopic values: replace 0-values in Kato and in urine with new minimum values.

worms.sub <- subset(worms$Kato, worms$Kato > 0)
y <- min(worms.sub, na.rm = TRUE)
## [1] 20

KatoMn <- y/4
## [1] 5

worms$Kato <- replace(worms$Kato, worms$Kato == 0, KatoMn)

worms.sub <- subset(worms$urine, worms$urine > 0)
z <- min(worms.sub, na.rm = TRUE)
## [1] 0.5

urineMn <- z/10
## [1] 0.05

worms$urine <- replace(worms$urine, worms$urine == 0, urineMn)

Check the new correlations. Now it looks much better.

cor(worms$Ct_Sm, worms$Kato)
## [1] -0.3707
cor(worms$Ct_Sh, worms$urine)
## [1] -0.3315

And let us make some nice plots.

p <- qplot(worms$Ct_Sm, log(worms$Kato), worms)
p + geom_smooth(method = "lm")

plot of chunk unnamed-chunk-6

w <- qplot(worms$Ct_Sh, log(worms$urine), worms)
w + geom_smooth(method = "lm")

plot of chunk unnamed-chunk-6



Thesis: Molecular Detection of Intestinal Parasites for Clinical Diagnosis and Epidemiology



The detection of intestinal parasitic infections for routine diagnosis and for epidemiological research still depends mainly on microscopical examination of stool samples for the identification of helminth eggs and protozoan trophozoites and cysts. Because microscopy has several limitations, additional diagnostic methods (e.g. culture, antigen/antibody detection) have been accessed to surpass obstacles in detection and characterization of intestinal parasites. Although such additional methods increases sensitivity, the amount of hands-on time accumulates substantially.

During the last years remarkable progress has been made on another diagnostic methods that are based on Polymerase Chain Reaction (PCR) technique. DNA isolation from stool can be processed in a semi- or fully-automated system where after specific DNA of multiple targets can be simultaneously amplified, visualized and semi-quantified in a closed tube system with multiplex real-time PCR. The molecular diagnostic approach was merged with an alternative diagnostic strategy where clusters of patients with shared characteristics are routinely screened for a selected number of parasites species. This new diagnostic strategy was assessed for the routine diagnosis and epidemiology of intestinal parasites in patients consulting general practitioners.

Samples of a general practice patient group were processed on a real-time PCR panel for the detection of Entamoeba histolytica, Giardia lamblia and Cryptosporidium (HGC PCR). The retrospective analysis with HGC PCR was compared with the results obtained with those of routinely performed microscopy. The results revealed that significant numbers of G. lamblia and Cryptosporidium infections remained undetected with microscopy. On the other hand, the parasites that have been detected with microscopy but not with real-time PCR consisted mainly of non-pathogenic parasites and Dientamoeba fragilis, although the pathogenicity of the latter is disputed. The results showed that, compared to the molecular approach, microscopy provided limited additional value in routine diagnosis of general practice patients with gastro-intestinal complaints, even with the use of multiple sampling procedure of faecal specimens in combination with fixatives (also referred to as the triple faeces test (TFT) procedure). The introduction of the TFT method in Dutch laboratories resulted of D. fragilis being increasingly diagnosed. Although D. fragilis has been suggested to be a potential pathogen in children, more studies are needed to support this statement. The introduction of a D. fragilis real-time PCR in routine diagnostics can help to elucidate the pathogenicity of this parasite. Cases of clinical E. histolytica infections (amoebiasis) in The Netherlands are very rare and can be associated to travel in areas where the parasite is endemic. Still, in several cases the source of infection could not be explained. Unselected screening for this parasite can be appreciated because of its major clinical importance for the patient and its potential to spread among household members.

Many intestinal parasites species, such as hookworms or Cyclospora cayetanensis, have adapted their transmission pattern and life cycle to specific environmental factors. Being bound to these factors, these parasites appear only occasionally in The Netherlands and surrounding countries; usually they are taken along by travellers returning from countries where the parasites are endemic. Real time PCRs for more exotic parasites can be used as extensions of basic molecular assays in clinical diagnostic settings. The relative frequency of a wide variety of intestinal parasitic diseases has been assessed in a population of travellers (such as tourists, immigrants, expats, etc.) using two different diagnostic approaches. The “overall view” of the conventional approach (i.e. microscopy combined with antigen tests) was compared with the molecular diagnostic approach, where only those species that are targeted in the assay are detected. The molecular diagnostic method showed more effective in detecting the targeted parasites than the conventional diagnostic approach, in particular E. histolytica and Strongyloides stercoralis for which additional diagnostic techniques are needed op top of microscopy. Only few additional parasite species have been detected with microscopy, while many more unexplained causes of gastro-intestinal complaints might be resolved by the application of additional real-time PCR targets.

Using the molecular hight throughput screening system in combination with a simple collection procedure for stool specimens, valuable data has been produced that has yielded new insights in the epidemiology of several parasitic diseases.

The prevalence of Cryptosporidium in Dutch patients with gastro-intestinal complaints attending their general practitioner has so far exceeded the figures in previous studies. The highest infection rate has been detected among children aged under five years with a peak in the month of September; almost one-third of them had been infected with Cryptosporidium. Microscopic examination for Cryptosporidium, which requires an additional staining procedure of the faecal smear, was specifically requested by the general practitioner twenty-one times and was found positive in 13 cases once, whereas with HGC PCR 80 cases were detected. The lack of request for additional diagnostic procedures is not the only reason for missing infections. Basic microscopic stool examination has often not been requested, leaving a substantial number of gastro-intestinal parasitic infections undiagnosed.

Symptoms of G. lamblia infections are highly variable, and can range from asymptomatic to the presence of severe gastro-intestinal complaints. Data in this thesis showed that, although the G. lamblia Cycle-threshold (Ct)-values correlated with the number of specified gastro-intestinal complaints, the parasite was still detected in 4.7% of persons who did not have complaints. The reasons for the clinical heterogeneity of G. lamblia infections are not fully understood. Although the epidemiological role of G. lamblia assemblages (i.e. group of genotypes) had been suggested as an important factor associated with gastro-intestinal complaints, this could not be proven for adults. Nevertheless, in children the assemblages might still be relevant for the clinical presentation of the patient.

In this thesis, additional real-time PCRs have been evaluated to cover the most important intestinal parasite species for epidemiological studies and to be used as extensions of basic molecular diagnostic assays in clinical settings. Isospora belli (recently renamed as Cystoisospora belli), Encephalitozoon intestinalis and Enterocytozoon bieneusi are opportunistic pathogens that can cause life-threatening diarrhoea and malabsorption, in immuno-compromised patients. With increased awareness, infections with E. bieneusi are diagnosed not only in immune-compromised patients, but also showed presence in asymptomatic persons. Furthermore, the high prevalence of weak infections or low shedding of spores (high Ct-values) have been observed among populations in sub-Saharan- and East-African countries. The phylogenetic study on microsporidia infections in persons with different clinical backgrounds indicated a dynamic evolutionary process between genotypes of E. bieneusi. One specific genotype was restricted to transplantation patients receiving immuno-supressives and another genotype showed its preferential habitat in patients living with HIV/AIDS, which further emphasizes the predisposition for specific hosts by different E. bieneusi isolates.

Last but not least, real-time PCRs were developed for the detection of Schistosoma mansoni and Schistosoma haematobium. Both Schistosoma species could successfully be detected in stool samples with Ct-values correlating with the results of quantitative microscopy. S. haematobium could also be detected in DNA isolated both from urine and from cervico-vaginal lavages. Results indicated that real-time PCR may potentially serve as a gold standard to determine the prevalence and intensity of Schistosoma infections in surveys. The real-time PCR may even be provided as a diagnostic tool for urine, stool and other clinical samples. This might prove helpful in particular for the difficult to diagnose female genital schistosomiasis. The semi-quantitative outcome of the PCR might be used as a predictor in the disease pathology and help to determine if observed lesions are caused by the parasite eggs.

Full-text PDF

Chapter 9 – Multiplex real-time PCR for the detection and quantification of Schistosoma mansoni and S. haematobium infection in stool samples collected in northern Senegal


A multiplex real-time PCR assay for the detection and quantification of Schistosoma mansoni and S. haematobium DNA in faecal samples was developed and evaluated as an alternative diagnostic method to study the epidemiology of schistosomiasis. Primers and probes targeting the cytochrome c oxidase gene were designed for species-specific amplification and were combined with an internal control. Using positive control DNA extracted from adult Schistosoma worms and negative control samples (n = 150) with DNA from a wide range of intestinal microorganisms, the method proved to be sensitive and 100% specific. For further evaluation, duplicate stool specimens with varying S. mansoni egg loads were collected in northern Senegal from pre-selected individuals (n = 88). The PCR cycle threshold values, reflecting parasite-specific DNA loads in faeces, showed significant correlation with microscopic egg counts both for S. mansoni in stool and S. haematobium in urine. The Schistosoma detection rate of PCR (84.1%) was similar to that of microscopy performed on duplicate stool samples (79.5%). The simple faecal sample collection procedure and the high throughput potential of the multiplex real-time PCR provide a powerful diagnostic tool for epidemiological studies on schistosomiasis in remote areas, with possibilities for extension to other helminths or protozoa using additional molecular targets.

Full-text PDF

R.J. ten Hove, J.J. Verweij, K. Vereecken, K. Polman, L. Dieye, L. van Lieshout. Transactions of the Royal Society of Tropical Medicine and Hygiene (2008), 102(2): 179-185