Archive for Blog

Profiling of Antibiotic Resistance of Bacterial Species Recovered from Routine Clinical Isolates in Ethiopia.

Melaku Tesfaye1
Witold Frederik ten Hove2
Mesfin Nigussie1
1. International Clinical Laboratories, Addis Ababa, Ethiopia
2. HAN University of Applied Sciences, Nijmegen, the Netherlands

This is a guide to explain the aggregation and visualisation of raw data extracted from the Polytech Laboratory Information Management System, at International Clinical Laboratories, Addis Ababa, Ethiopia. The guide starts with a piece of R-code, followed by an explanation.


# libraries

The working directory need to be set and, I like to put all libraries together.

# Read text file (file should be in the same working directory as this script)
con <- file("antibiogram_raw.txt")
txtData <- readLines(con, warn = FALSE, skipNul = TRUE)
## [1] "Page 1/176\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                            
## [2] "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                      
## [3] "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                      
## [4] "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                      
## [5] "Sensitive/Resistant/Intermediate Count\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"
## [6] "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"

The raw data was exported by a query from LIMS. It resulted in 176 PDF pages. With Copy-Paste, the data was then saved as a simple .txt file. This text file is read into R. First as a file(con) and then as a string (txtData).

# Clean up
txtData <- gsub("\t{1,}", ";", txtData)
txtData <- gsub("^Page", NA, txtData)
txtData <- gsub("Site:", "", txtData)
txtData <- gsub("^;$", NA, txtData)
txtData <- gsub("^Sensitive.*", "New", txtData)
txtData <- gsub("^;", "", txtData)
txtData <- gsub(";$", "", txtData)
txtData <- gsub("S;R;I", "Data", txtData)
txtData <- na.omit(txtData)

With the gsub command, specific texts can be replaced (also by ‘nothing’:“”). The function na.omit(txtData) returns the object txtData with listwise deletion of missing values.

# Rename Antibiotics
txtData <- gsub("^Ampic[a-z]{1,}", "AMP10", txtData)        
txtData <- gsub("^Amox/K Clav'ate", "AMC30", txtData)   
txtData <- gsub("^Chloramp[a-z]{1,}", "C30", txtData) 
txtData <- gsub("^Ceftriax[a-z]{1,}", "CRO30", txtData) 
txtData <- gsub("^Cephaloph[a-z]{1,}", "CF30", txtData) #Cephalophin doesn't exist
txtData <- gsub("^Cephaloth[a-z]{1,}", "CF30", txtData) 
txtData <- gsub("^Ciproflo[a-z]{1,}", "CIP5", txtData) 
txtData <- gsub("^Clindam[a-z]{1,}", "CC2", txtData) 
txtData <- gsub("^Cefotax[a-z]{1,}", "CTX30", txtData) 
txtData <- gsub("^Cefurox[a-z]{1,}", "CXM30", txtData) 
txtData <- gsub("^Cefoxit[a-z]{1,}", "FOX30", txtData) 
txtData <- gsub("^Erythrom[a-z]{1,}", "E15", txtData) 
txtData <- gsub("^Gentam[a-z]{1,}", "GM10", txtData) 
txtData <- gsub("^Metroni[a-z]{1,}", "METR", txtData)
txtData <- gsub("^Nalidixic Acid", "NA30", txtData) 
txtData <- gsub("^Norfloxa[a-z]{1,}", "NOR10", txtData) 
txtData <- gsub("^Nitrofur[a-z]{1,}", "FM300", txtData) 
txtData <- gsub("^Oxac[a-z]{1,}", "OX1", txtData) 
txtData <- gsub("^Penic[a-z]{1,}", "P10", txtData) 
txtData <- gsub("^Pyrazinamide{1,}", "PYR", txtData)
txtData <- gsub("^Sulfameth[a-z]{1,}", "SXT25", txtData) 
txtData <- gsub("^Tetrac[a-z]{1,}", "TE30", txtData) 
txtData <- gsub("^Vancom[a-z]{1,}", "VA30", txtData) 
txtData <- gsub("^Imip[a-z]{1,}", "IMP", txtData) 
txtData <- gsub("^Rifamp[a-z]{1,}", "RIF", txtData)  
txtData <- gsub("^Ethamb[a-z]{1,}", "ETB", txtData) 
txtData <- gsub("^Streptom[a-z]{1,}", "STREP", txtData) 

Antibiotics are renamed into easy to use abbreviations. Also, often the antibiotics are mispelled, e.a. Ampicilline / Ampiciline / Ampicyline…

# Function to split vector into list
VectorToList <- function(vec, ind) {
  n = 1
  splitIndex <- which(vec == ind)
  myList <- NULL
  for (i in splitIndex[-length(splitIndex)]) {
    myList[[n]] <- vec[(i+1):(splitIndex[n+1]-1)]
    n <- n+1

Prepare a function which will split the vector (txtData) into a list (myList).

# Function to split vector into data frame
VectorToDF <- function(vec, ind){
  d <- which(vec[[1]] == ind)+1
  l <- length(vec[[1]])
  myDf <- data.frame(
        vec[[1]][d:l], split = ";", fixed = FALSE, perl = FALSE, useBytes = FALSE
    , ncol = 4, byrow = TRUE

The next function is to split the vector and make a dataframe (myDF) from it.

SelectListItem <- function(lst, elnt = 1, lkp) {
  itemVec <- sapply(lst, function(x) x[[elnt]][[1]]) # Create vector with all items from selected list element
  myList <- lst[which(itemVec == lkp)] # Use which function on vector to create vector of indexnumbers and select items from original list

Function to select elements from list.

# Create list of separate vectors
txtList <- VectorToList(txtData, "New") 

‘txtList’ is created with comprehensable chunks of data (lists).

# Create list of lists with results in data frame
newList <- NULL
for (m in 1:length(txtList)) {
  testresDf <- VectorToDF(txtList[m], "Data")
  colnames(testresDf) <- c("Antibiotics", "S", "R", "I")
  newList[[m]] <- list(bacteria = txtList[[m]][1], source = txtList[[m]][2], result = testresDf)

The function VectorToDF aggregates txtList into Dataframe

# Create data frame from list
newDF <- NULL
for (s in 1:length(newList)) {
  cntRows <- length(newList[[s]]$result[,1])
  bact <- rep_len(newList[[s]]$bacteria,cntRows)
  srce <-  rep_len(newList[[s]]$source,cntRows)
  tmpDF <- cbind(Bacteria = bact, Source = srce, newList[[s]]$result)
  newDF <- rbind(newDF, tmpDF)
newDF$Bacteria <- gsub("(^|[[:space:]])([[:alpha:]])", "\\1\\U\\2", newDF$Bacteria, perl=TRUE) # Change all first letters of bacteria names to upper case

Now to make a workable database.

# Correct Bacteria species
newDF$Bacteria <- gsub("?Species", "species", newDF$Bacteria) 
newDF$Bacteria <- gsub("Actinomyces Viscosus", "Actinomyces viscosus", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Fragilis", "fragilis", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Braakii", "braakii", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Farmeri", "farmeri", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Freundii", "freundii", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Koseri", "koseri", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Cloacae", "cloacae", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Faecalis", "faecalis", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Coli", "coli", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Influenzae", "influenzae", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Biotype", "biotype", newDF$Bacteria) 
newDF$Bacteria <- gsub("HEMOPHILUS INFLUENZA", "Haemophilus influenzae", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Ornithinolytica", "ornithinolytica", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Oxytoca", "oxytoca", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Pneumoniae", "pneumoniae", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Terrigena", "terrigena", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Lacunata", "lacunata", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Morganii", "morganii", newDF$Bacteria) 
newDF$Bacteria <- gsub("Mycobacteria Tuberculosis Conplex", "Mycobacterium tuberculosis complex", newDF$Bacteria)
newDF$Bacteria <- gsub("Mycobacterium Tuberculosis Complex", "Mycobacterium tuberculosis complex",newDF$Bacteria)
newDF$Bacteria <- gsub("?Mirabilis", "mirabilis", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Penneri", "penneri", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Vulgaris", "vulgaris", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Aeruginosa", "aeruginosa", newDF$Bacteria) 
newDF$Bacteria <- gsub("SALMONELLA GROUP.*", "Salmonella group", newDF$Bacteria) 
newDF$Bacteria <- gsub("?Paratyphi", "paratyphi", newDF$Bacteria) 
newDF$Bacteria <- gsub("?SALMONELLA PARATYPHI", "Salmonella paratyphi", newDF$Bacteria) 
newDF$Bacteria <- gsub("Boydii", "boydii", newDF$Bacteria)
newDF$Bacteria <- gsub("Dysenteriae", "dysenteriae", newDF$Bacteria) 
newDF$Bacteria <- gsub("SHIGELLA GROUP", "Shigella group", newDF$Bacteria) 
newDF$Bacteria <- gsub("Sonnei", "sonnei", newDF$Bacteria)
newDF$Bacteria <- gsub("Aureus", "aureus", newDF$Bacteria) 
newDF$Bacteria <- gsub("Lugdunensis", "lugdunensis", newDF$Bacteria) 
newDF$Bacteria <- gsub("Saprophyticus", "saprophyticus", newDF$Bacteria)
newDF$Bacteria <- gsub("Agalactiae", "agalactiae", newDF$Bacteria) 
newDF$Bacteria <- gsub("Streptococcus Anginosus/milleri", "Streptococcus anginosus/milleri", newDF$Bacteria)
newDF$Bacteria <- gsub("Pyogenes", "pyogenes", newDF$Bacteria)
newDF$Bacteria <- gsub("4\\(\\ Erwina species)", "4 (Erwina species)", newDF$Bacteria) 
newDF$Bacteria <- gsub("Viridans Streptococcus", "Viridans streptococci", newDF$Bacteria)
newDF$Bacteria <- gsub("Coagulase-negative Staphylococcus species", "CNS", newDF$Bacteria)

The bacteria species names contain many mistakes. They need to be corrected one by one.

## change factors into numeric
newDF$S <- as.numeric(newDF$S)
newDF$I <- as.numeric(newDF$I)
newDF$R <- as.numeric(newDF$R)
## summ all cultures (Sensitive + Intermediate + Resitant)
newDF$N <- newDF$S + newDF$I + newDF$R

The numbers under S, I & R are stored as ‘factors’. They can be changed into numeric with command ‘as.numeric’. ‘N’ is the sum off antibiograms.

###  Tuberculosis and its drugs
myco <- newDF[grep("Mycobacterium tuberculosis complex", newDF$Bacteria), ]
## [1] 334
write.csv(myco, "myco.csv")
newDF <- newDF[-grep("Mycobacterium tuberculosis complex", newDF$Bacteria), ]
## [1] 15989

First make a seperate dataframe for Mycobacteria species. Next is to remove Mycobacterium from the dataframe by using the command ‘-grep’. The commands ‘sum’ show that 334 Mycobacterium antibiograms were removed, leaving 15989 remaining antibiograms.

# Cleaning up Source
## Rename 'W' and 'WOUND' as 'Wound' 
newDF$Source <- gsub("WOUND", "AIAIAI", fixed = TRUE, newDF$Source) 
newDF$Source <- gsub("W", "Wound", fixed = TRUE, newDF$Source) 
newDF$Source <- gsub("AIAIAI", "Wound", fixed = TRUE, newDF$Source) 

newDF$Source <- gsub("BODY FLUID", "Body fluid", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("BLOOD-PEADITRIC", "Blood-pediatric", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("BLOOD-AEROBIC", "Blood-aerobic", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("BLOOD-ANAEROBIC", "Blood-anaerobic", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("EAR CULTURE", "Ear culture", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("NASAL DISCHARGE", "Nasal discharge", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("PUS", "Pus", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("ROUTINE", "Routine", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("SPUTUM", "Sputum", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("STOOL", "Stool", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("Thorat culture", "Throat culture", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("EYE CULTURE", "Eye culture", fixed = TRUE, newDF$Source)

Correct the ‘SOURCE’ and change to lower case

### Merge Blood-anaerobic with Blood-aerobic to simply 'Blood' 
newDF$Source <- gsub("Blood-aerobic", "Blood", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("Blood-anaerobic", "Blood", fixed = TRUE, newDF$Source)

### Merge "Throat culture" with "Throat swab" to simply "Throat"
newDF$Source <- gsub("Throat swab", "Throat", fixed = TRUE, newDF$Source)
newDF$Source <- gsub("Throat culture", "Throat", fixed = TRUE, newDF$Source)

Merging the ‘Blood-aerobic’ and ‘Blood-anaerobic into simply ’Blood’ The same for ‘Throat swab’ and ‘Throat culture’.

### Aggregate double sources and antibiotics.
newDF <- aggregate(cbind(S,I,R,N)~Bacteria+Source+Antibiotics, data=newDF, sum, na.rm=TRUE)
## [1] 15989

Many lines now contain the same Bacterium species with same source Source. These are grouped together with the R, S and I’s summed up.

## Export data
write.csv(newDF, "newDF.csv")

An intermediate raw-data file is made.

Bacteria Source Antibiotics S I R N
CNS Blood AMC30 11 3 6 20
Citrobacter braakii Blood AMC30 4 2 2 8
Enterobacter cloacae Blood AMC30 1 1 3 5
Escherichia coli Blood AMC30 7 14 21 42
Group C Streptococcus Blood AMC30 1 1 3 5
Klebsiella pneumoniae Blood AMC30 2 2 5 9
Pseudomonas species Blood AMC30 1 3 3 7
Staphylococcus aureus Blood AMC30 10 4 2 16
Staphylococcus lugdunensis Blood AMC30 4 2 6 12
Streptococcus pyogenes (Group A) Blood AMC30 4 2 2 8
Actinomyces viscosus Blood-pediatric AMC30 2 1 1 4
Bacteroides fragilis Blood-pediatric AMC30 2 1 1 4
CNS Blood-pediatric AMC30 8 1 1 10
Escherichia coli Blood-pediatric AMC30 3 2 3 8
Group A Streptococcus Blood-pediatric AMC30 2 1 1 4
Klebsiella oxytoca Blood-pediatric AMC30 2 2 6 10
Pseudomonas species Blood-pediatric AMC30 2 2 6 10
The dataframe (first 17 rows).
newDF2 <- read.csv("newDF.csv", sep = ",")
newDF2 <- subset(newDF, select=c(Bacteria, Source, Antibiotics, S, I, R, N))

To start with a fresh new Dataset

### Aggregate on Species.
DFspecies <- aggregate(cbind(N)~Bacteria, data=newDF2, sum, na.rm=TRUE)
Bacteria species with number of antibiograms
Bacteria N
Acinetobacter species 54
Actinomyces viscosus 37
Bacteroides fragilis 8
Beta Hemolytic Non-group A 30
CNS 2585
Citrobacter braakii 64
Citrobacter farmeri 33
Citrobacter freundii 39
Citrobacter koseri 40
Enterobacter cloacae 140
Enterobacter species 68
Enterococcus faecalis 619
Enterococcus species 50
Escherichia coli 2931
Group A Streptococcus 128
Group B Streptococcus 224
Group C Streptococcus 113
Group F Streptococcus 38
Group G Streptococcus 42
Haemophilus influenzae 47
To give an overview on the nu mber of antibiograms per Bacteria species (only first 20 rows).
##### Group species together
##### \\ = space    [a-z] = any character   {,1} = repeat previous 1 or more time
# put together Citrobacter species (C. braakii, farmeri, freundii & koseri)
newDF2$Bacteria <- gsub("Citrobacter\\s[a-z]{1,}", "Citrobacter spp.", newDF2$Bacteria) 
# put together Enterobacter cloacae & species
newDF2$Bacteria <- gsub("Enterobacter\\s[a-z]{1,}", "Enterobacter spp.", newDF2$Bacteria) 
# put together Enterococcus faecalis & species
newDF2$Bacteria <- gsub("Enterococcus\\s[a-z]{1,}", "Enterococcus spp.", newDF2$Bacteria) 
# put together Klebsiella ornithinolytica, oxytoca, pneumoniae, terrigena & species
newDF2$Bacteria <- gsub("Klebsiella\\s[a-z]{1,}", "Klebsiella spp.", newDF2$Bacteria) 
# put together Proteus mirabilis, penneri, vulgaris & species
newDF2$Bacteria <- gsub("Proteus\\s[a-z]{1,}", "Proteus spp.", newDF2$Bacteria) 
# put together Pseudomonas auruginosa & species
newDF2$Bacteria <- gsub("Pseudomonas\\s[a-z]{1,}", "Pseudomonas spp.", newDF2$Bacteria) 
# put together Salmonella group, paratyphi A & B & species
newDF2$Bacteria <- gsub("Salmonella\\s[a-z]{1,}", "Salmonella spp.", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Salmonella spp.\\s[AB]", "Salmonella spp.", newDF2$Bacteria) 
# put together Shigella boydii, dysenteriae, group A1, sonnei & species
newDF2$Bacteria <- gsub("Shigella\\s[a-z]{1,}", "Shigella spp.", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Shigella spp.\\sA1", "Shigella spp.", newDF2$Bacteria)
# put together Staphylococcus lugdunensis & saprophyticus as CNS (keep S. aureus seperate)
newDF2$Bacteria <- gsub("Staphylococcus lugdunensis", "CNS", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Staphylococcus saprophyticus", "CNS", newDF2$Bacteria)
# Rename Group [A-Z] Streptococcus as 'Streptococcus Group [A-Z]'
newDF2$Bacteria <- gsub("Group\\sA\\sStreptococcus", "Streptococcus Group A", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Group\\sB\\sStreptococcus", "Streptococcus Group B", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Group\\sC\\sStreptococcus", "Streptococcus Group C", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Group\\sF\\sStreptococcus", "Streptococcus Group F", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Group\\sG\\sStreptococcus", "Streptococcus Group G", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Streptococcus\\sagalactiae\\s[:(:]Group\\sB[:):]", "Streptococcus Group B", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Streptococcus\\sanginosus[:/:]milleri", "Streptococcus Group F", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Streptococcus\\spyogenes\\s[:(:]Group\\sA[:):]", "Streptococcus Group A", newDF2$Bacteria)
# Put together Streptococcus species & Viridans
newDF2$Bacteria <- gsub("Streptococcus\\sspecies", "Streptococcus spp.", newDF2$Bacteria) 
newDF2$Bacteria <- gsub("Viridans\\sstreptococci", "Streptococcus spp.", newDF2$Bacteria) 
# In case needed, put all Streptococcus Group [A-Z] together as Beta-hemolytic Streptococci
newDF2$Bacteria <- gsub("Streptococcus\\sGroup\\s[A-Z]", "Beta-hemolytic Streptococci", newDF2$Bacteria) 

Instead of having small groups of Bacteria species, they are merged together into bacteria genus.

# aggregate new groups.
newDF2 <- aggregate(newDF2[,c("S","I","R", "N")], by=list(newDF2$Bacteria, newDF2$Source, newDF2$Antibiotics), "sum")
names(newDF2)[names(newDF2)=="Group.1"] <- "Bacteria"
names(newDF2)[names(newDF2)=="Group.2"] <- "Source"
names(newDF2)[names(newDF2)=="Group.3"] <- "Antibiotics"

The same Bacteria genus rows are merged together with aggregate comment. New names are put in the header.

newDF2$n <- (newDF2$S + newDF2$R)

The sum is calculated for sensitive plus resistant antibiograms.

# When calculating %resistance, leave out the intermediates. 
newDF2$Resis <- ((newDF2$R/(newDF2$S + newDF2$R))*100)
newDF2$Resis <- formatC(newDF2$Resis, digits=3)
newDF2$Resis <- as.numeric(newDF2$Resis)
#sumDF$n <- (sumDF$S + sumDF$R)
## [1] 15989
## [1] 13073

For this study, the resistance rate is calculated from the sum of sensitive plus resistant cultures, leaving the intermediate out.

newDF3 <- newDF2[newDF2$n > 6,]

After keeping out the intermediates, also groups are left out that contain only 6 or less cultures. Setting the threshold on 6 is on one hand ambiguous. However, after checking all the group-size, it was observed that a particular large proportion of ‘small groups’ was removed when setting the threshold on 7 or more cultures per group.

### new dataset without SOURCE, aggregated again ####
Plot7 <- aggregate(newDF3[,c("S","R","n")], by=list(newDF3$Bacteria, newDF3$Antibiotics), "sum")
names(Plot7)[names(Plot7)=="Group.1"] <- "Bacteria"
names(Plot7)[names(Plot7)=="Group.2"] <- "Antibiotics"

Plot7$Resis <- ((Plot7$R/Plot7$n)*100)

Plot7$Resis <- formatC(Plot7$Resis, digits=3)
Plot7$Resis <- as.numeric(Plot7$Resis)

## [1] 10684
## [1] 10684

Now, the dataset is aggregated with antibiotics on the Y-axis and bacteria genus on the x-axis.


Plot7g <- ggplot(Plot7) + geom_point(aes(x = Bacteria, y = Antibiotics, colour = Resis, size = n)) +
  scale_colour_gradient(name = '% resistance', high = "red", low = "green") +
  scale_size(name = 'sample size', range = c(2,9)) +
  xlab('Bacteria groups') + ylab('Antibiotic') +
  theme(text = element_text(size = 12), axis.text.x = element_text(angle=60, hjust=1, size=12) )


And plotting the aggregated dataset.


A month ago, during an interview for a new position at the European Commission, I was asked if I could deal with the infamous bureaucracy. I looked at them… and laughed… and laughed. Guys, I work in Ethiopia!

So now, I sold my car and the ownership needs to be transferred. A two minute procedure in most countries. In Ethiopia, day two and I’m still stuck in bureaucratic madness. I feel like I’m in Franz Kafka’s wet dream.

Step one.

Drive to the road authorities where the car is registered. Luckily, it’s in Addis Ababa and not in Axum or something. At the entrance, together with the new owner and three witnesses, I need to fill in this selling-agreement document. Later, the tax is calculated over the selling price. The selling price is below a certain threshold, therefor, the authorities fixed a selling price for us.

The witnesses can be arranged on the spot for a small compensation. If I want to bypass the lines at the road authorities, a witness said, he could arrange it for 1000,- etb. Not negotiable because many people then need to be paid. I refused as I came prepared with a very thick novel.

After a couple of hours in line, I was summoned to a window. A minute later I was told to come back tomorrow. Closing time…

Step two

The selling of the car needs to be authenticated by the court at a notary’s office, or something. I am married and therefore I’m fucked. Literally and figuratively.

Step … oh fuck it…

  • My wife has to be at the notary in person. Check.
  • I need to present a marriage certificate. Check
  • The certificate needs to be in English. Check
  • The certificate needs to be stamped and signed by the Netherlands ministry of foreign affairs in the Hague. Check
  • The certificate needs to be stamped and signed by the Ethiopian embassy in Brussel. Check
  • The certificate needs to be stamped by the Ethiopian ministry of foreign affairs. Check
  • The Ethiopian stamp was older then 6 months, so i needed to get a new stamp. oookaaaay
  • The certificate was just over 2 years old SO THEY REFUSED TO STAMP IT…

…aaaand the usual YOU GO AWAY! Yes, I am trying. I am trying.

to be continued.


This document is dedicated to my old-uncle Stanisław Czernicki.dyrektor-stanislaw-czernicki

Recently, I found a Polish book: ‘Stories from teachers’. The first chapter was written by Stanisław and tells the story of his escape from the concentration camp. This project is to translate the story from Polish to English, and to do a background research on the mentioned names.

Here, the first page. The full story can be read on my GitHub website.

    Source of photo

Read more

Genotyping Hepatitis C Virus

The Hepatitis C virus (HCV) is a single-stranded RNA virus, with a genome of 9,500 nucleotides. The prevalence of HCV among the general population of Ethiopia is not known. Nonetheless, the impression at the lab is that prevalence is high. Egypt is know to have one of the highest HCV prevalence in the world. It is postulated that this is the result of mass treatment with praziquantel for schistosomiasis using contaminated syringes in the past. Mass treatment against schistosomiasis also occurred in Ethiopia, however, oral treatment were used. Intravenous drug users are also unknown of. Would it be facial tattoos, high number of cesarians, blood transfusions? The primarily mode of transmission is still a mystery.

About 75 – 85% of HCV-infected individuals develop chronic hepatitis, with up to 20% of chronically infected individuals developing cirrhosis. In patients with cirrhosis, the incidence of hepatocellular carcinoma is 1 – 4% per year. Genotype 1 and genotype 4 responds less well to interferon-based anti-HCV drug treatment than genotypes 2 and 3, especially in patients of African descent. Therefore, genotyping of HCV is recommended before the start of antiviral therapy.

Thanks to the lobbying of, among other PharmAccess, treatment for HCV is becoming more affordable in developing countries. Still, hepatitis C viral loads and genotyping is mandatory for managing the treatment.

The Abbott RealTime HCV genotype II assay is an RT-PCR assay for use with the Abbott mSample Preparation System reagents and with the Abbott m2000sp and m2000rt instruments for viral count and genotyping of the viral (HCV) RNA in human blood, serum plasma (EDTA) or tissue donors of HCV-infected individuals. The Abbott RealTime HCV assay is intended for use as an aid in the management of HCV-infected patients undergoing antiviral therapy. The assay measures HCV RNA levels at baseline and during treatment and can be utilized to predict sustained and non-sustained virological response to HCV therapy.

During the verification tests, some remarkable observations were made. There was failure of the internal controls and mismatch at low viral loads. But that’s not the interesting one.

The genotype 4e at the reference laboratory in Germany (most likely they use the Roche system) corresponds with a genotype 1 & 4 double reaction at our lab in Addis Ababa. Genotype 4e is a strain of which little is known. There are only few and partial RNA sequences available for this genotype. It is therefore hypothesized that the primers/probes for gt1 detection from the Abbott assay cross match with genotype 4e.

HCV genotyping


In case, with the Abbott system, a reaction occurs in which both genotype 1 & 4 are detected, the results should be interpreted as HCV genotype 4.


Power analysis in R

Power analysis