2. 1) Early diagnosis of cancer
Cancer is one of the widespread diseases leading to fatal death
Among them, lung cancer and breast cancer accounts the most
It has been found that if the disease is being diagnosed at an early stage,
the survival rate of the patient could be improved
However, in most of cases, the disease is being diagnosed at a later stage
3. 1) Early diagnosis of cancer
There is an efficacious screening method (high resolution mammography),
to be used for screening target female population on breast cancer
The similar method does not exist for early detection of lung cancer
Hence, the early diagnosis of lung cancer is the right topic to be an
objective of predictive modelling, performed within DM and ML methods
4. Rajan JR, Prakash JJ. Early diagnosis of lung cancer using a
mining tool. IJETICS, ISSN 2278-6856, 2013.
The hospital reports of the patients suffering from lung cancer have been
collected from various sources and integrated by the classification
technique based on using ANN (Arteficial Neural Networks).
The expert system, that was defined this way to support the diagnosis of
lung cancer, was based on using information on risk factors and symptoms,
rather than on images or biomarkers.
Such approach is especially appropriate for use in less developed countries
(India as an example).
5. 2) Predicting response to a drug therapy for cancer
Cancer is a clinically and genomically heterogeneous disease
Pharmaceutical companies apply DM to huge masses of genomic data to
predict how a patient`s genetic makeup determines his/her response to a
drug therapy
6. The Use Case: Daemen A et al. Modelling precision
treatment of breast cancer. Genome Biology 2013; 14:R110
First-generation molecular profiles for human breast cancer have enabled
identification of features that can predict therapeutic response
However, little is known about how the various data types can best be
combined to yield optimal predictors
In this paper, collections of breast cancer cell lines was used to set cancer
molecular pathobiology (omics- measurements) with biological therapeutic
responses, in order to identify the most predictive molecular feature sets
7. The Use Case: Daemen A et al. Modelling precision treatment of breast
cancer. Genome Biology 2013; 14:R110
Molecular profiling data sets included: profiles for DNA copy number,
mRNA expression, transcriptome sequence, promoter methylation, protein
abundance and mutation status
Least squares-support vector machines and random forest algorithms were
used to identify molecular features associated with responses
8. The Use Case: Daemen A et al. Modelling precision treatment of breast
cancer. Genome Biology 2013; 14:R110
Results showed that matching patients to treatments based on
transcriptional subtype will improve response rates, but that inclusion of
additional features (from other profiling data types) may provide additional
benefit
A systems biology strategy has been suggested to guide clinical trials, that
patient cohorts most likely to respond to new therapies may be more
efficiently identified
9. The Use Case: Daemen A et al. Modelling precision treatment of
breast cancer. Genome Biology 2013; 14:R110 (Fig.1)
Fig.1. Molecular profilling procedures of the panel of breast cancer cell lines
10. 3) Using image diagnostics and data mining tools for early
detection of other important chronic diseases
The art of modern medicine is to identify subjects with early stages of chronic
diseases, to enable preventive measures planning and timely treatment
A lot of money has been spent in biomarkers development
In addition to biochemical and molecular biology biomarkers, advanced
imaging techniques have been emerged as a source of biomarkers
11. Dai Z, He Y. Disrupted structural and functional brain connectomes in mild
cognitive impairment and Alzheimer`s disease. Neurosci Bull 2014; 30(2): 217–
232.
Efforts of researchers, in this study, have been oriented towards using standard diagnostic
imaging, to diagnose Alzheimer`s disease (AD), especially in its very early stages
By using graph theory analyses, this study showed that AD disease and its early phases of
development exhibit disrupted topological organization of large-scale brain networks and
that this loss correlates well with the decline in cognitive functions
By using combination of graph theory analyses and visual computer-based techniques, to
enable user-friendly interfaces –
very complex disease descriptions, based on using imaging diagnostic techniques,
can become standard output reports on the desks of medical doctors
and improve their insights into disease pathophysiology
12. 4) Using omics-techniques to improve understanding of
pathophysiology of chronic diseases
The enormous progress in biotechnology and computer-based techniques
for massive data analysis, in the last decades,
has provided clear benefits in medical practice
by implementing genomics, proteomics and other OMICS-techniques in
many sectors of health care
By using these techniques, it is possible to diagnose many chronic diseases
much earlier than before
and to enable knowledge on molecular mechanisms and signaling
pathways of many disorders - that has not been possible before
The highest grade of these information integration has been achieved
within systems biology and complex network analysis, holistic approaches
aimed at revealing the function of biological regulatory networks
13. 4) Using omics-techniques to improve pathophysiology understanding
The human microbiome is composed of microbes (mostly bacteria), that
reside in the gut
These microbes have tremendous potential to impact our physiology, both
in health and in disease
They contribute metabolic functions, protect against pathogens, educate
the immune system, and, through these basic functions, affect directly or
indirectly most of our physiologic functions
14. 4) Using omics-techniques to improve pathophysiology understanding
The study of the human microbiome has been furthered by technological
advancements for performing culture-independent analyses
In most studies, the bacterial constituents of a microbial population are
identified by sequencing of the 16S rRNA-encoding gene followed by
comparison to known bacterial sequence databases
Metagenomic analysis by sequencing all microbial DNA in a complex
community has the additional advantage of assessing the genetic
potential of the microbial population
Other methodologies to analyze the microbial transcriptome, proteome,
and metabolome provide additional information at successive levels of
microbial physiology
15. The Use Case: Qin J, et al. A human microbial gene catalogue
established by metagenomic sequencing. Nature 2010; 464(7285): 59-
65.(Fig. 2)
Fig. 2. Bacterial species abundance differentiates patients with
inflammatory bowel disease (ulcerative collitis & Crohn`s disease)
and healthy individuals
16. 5) The vaccines development
At the end of the 20th century, most of the vaccines that have been based on using
traditional technologies (by killing and attenuating viruses causing the diseases) had been
developed
New technologies have been required to conquer the remaining pathogens, being
refractory to vaccine development
Remarkable progress was made during this period by the introduction of new technologies
such as recombinant DNA and chemical conjugation of proteins to polysaccharides, as well
as advances in the use of novel adjuvants
Additionally, a powerful tool came from the ability to access the genomes of
microorganisms
This technological revolution allowed for the first time the capacity to move beyond the rules
of Pasteur, using the computer to rationally design vaccines starting with information present
in the genome, without the need to grow the specific microorganisms
This new approach was denominated “reverse vaccinology”
17. The use case: Sette A, Rappuoli R. Reverse vaccinology: Developing
vaccines in the era of genomics. Immunity 2010; 33(4): 530-41. (Fig. 3)
18. The use case: Sette A, Rappuoli R. Reverse vaccinology: Developing vaccines in the era
of genomics. Immunity 2010; 33(4): 530-41.
Fig. 3 description
The first pathogen addressed by the reverse vaccinology approach was Meningococcus B (MenB)
This pathogen was refractory to vaccine development because its capsular polysaccharide is identical to a human self-
antigen, whereas the bacterial surface proteins are extremely variable
Many attempts to develop a vaccine using the traditional technologies – have failed
The project to sequence the MenB genome and to use the genomic information to develop a vaccine
Gene sequences were analyzed, and over 600 potential antigens were tested for antigenicity
Candidate sequences were expressed in Escherichia coli, and sera from immunized mice was obtained against each of
them
Analysis of the sera revealed more than 90 previously unknown surface located proteins (only 12 surface antigens were
known, of them, only 4-5 showed bacterial activity)
29 out of 90 discovered antigens, were able to induce antibodies that could kill the bacteria in vitro in the presence of the
complement
In subsequent years, the antigens inducing the best and broadest bactericidal activity, were selected and inserted into
prototype vaccines that were able to induce protective immunity against most of the MenB strains in mice
After successful preclinical studies, the MenB vaccine entered the long path of vaccine development
19. 6)Prediction of health outcomes by using time series analysis
A time serie is a serie of data points indexed in time order
Most commonly, a time serie is a sequence taken at successive equally
spaced points in time - thus, a sequence of discrete-time data
There are two main goals of time series analysis:
Recognition of patterns (trends) in some process evolution
Prediction of outcomes
20. The use case: Sacchi L, Dagliati A, Segagni D, Leporati P, Chiovato L, Bellazzi R.
Improving risk-stratification of Diabetes complications using temporal data mining. IEEE,
2015
Starting with the stratification of patients based on using temporal patterns
of CSA (the index of temporal Medication Acquisition)
Authors considered the clinical variables that characterize the patients’
clinical condition associated with the diagnosis of diabetes type 2
The stability in drug purchases turns out to be a marker for a group of
patients who have an overall more complex clinical situation
Chronic complications were shown to be more frequent in a group of
patients more prone for medication changes over time
21. The use case: Sacchi L, Dagliati A, Segagni D, Leporati P, Chiovato L, Bellazzi R.
Improving risk-stratification of Diabetes complications using temporal data
mining. IEEE, 2015 (Fig. 4)
Fig. 4. Predicting chronic complications in diabetic patients
based on time series analysis of medication purchase