imagine (Institute of molecular genetics and genetic engineering)
Not a member yet
3088 research outputs found
Sort by
A biotechnological process for the production of pyocyanin and 1-hydroxyphenazine using waste streams from the potato chips industry
Get PDFPotato industry is one of the major food industries that generates considerable amounts of
potato peels and wastewater (1). These wastes can then be utilized as components of
microbial media in biotechnological production of pyocyanin (PYO) and 1-hydroxyphenazine
(1-HP) using Pseudomonas aeruginosa (2). PYO and 1-HP possess important biological
activities, thus could be applied in the field of medicine, and can be used as biocontrolling
agents (2). However, their application is hindered due to high costs associated with their large
scale production.
In this work, we established a fermentation process which utilizes either potato peels or
potato wastewater as the sole nutrient source to obtain PYO and 1-HP. P. aeruginosa BK25H
strain was selected from our in-house collection. This approach afforded 10 mg/l PYO and
9 mg/l 1-HP using potato wastewater and 15 mg/l PYO and 11 g/l 1-HP using potato peels
after 24 h incubation. This work is the step towards zero-pollution and conversion of waste
to valuable microbial products.Biotechnology for a circular bioeconomy: 28 -29 march 2023. AFOB-EFB Virtual conferenc
Developing bioinformatics pipeline for processing environmental DNA metabarcoding sequencing data
Get PDFEnvironmental DNA (eDNA) is DNA present in an environmental sample, originating from
any biological material released from organisms living in that environment. This DNA can
be isolated, amplified, sequenced, and analyzed in order to examine the taxonomic richness
and abundance of different organism groups in the targeted environment. Methods of
eDNA metabarcoding thus offer a unique opportunity to systematically streamline and
scale-up regular biological assessments across many different environments of interest.
Recently, as a part of the project funded by European structural and investment funds,
Labena d.o.o. company established a modern laboratory in Zagreb focused on the research
and provision of services in the field of eDNA. In collaboration with the Institute Ruđer
Bošković we have been working on developing tests for analysis of water quality based on
the eDNA and, as part of the standardization and optimization of sample-to-results eDNA
analysis process, we developed a custom bioinformatics pipeline to facilitate efficient and
effective eDNA sequencing data analysis.
The pipeline was was written in Bash and utilizes several different algorithms to filter,
trim, merge, denoise and classify targeted eDNA sequences. Python-based scripts which
allow automatically download, filter, and format the data available on various online
platforms were included in the pipeline to facilitate the curation of custom reference
databases needed for taxonomic classification of targeted organism groups. User-friendly
and interactive pipeline report generation, comprised of both wet- and dry-lab step-bystep
sample statistics and graphical representations or the main results, is supported
using Rmarkdown and Plotly and DataTables libraries. The pipeline is containerized in
Docker, allowing for easier environment building and pipeline deployment.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
PACSIN2 modifies miRNAs in extracellular vesicles, modulating thiopurine response
Get PDFThiopurines, such as mercaptopurine, are antimetabolites, used in the treatment of acute
lymphoblastic leukemia (ALL) and inflammatory bowel disease (IBD). PACSIN2 rs2413739
is associated with gastrointestinal toxicity in children with ALL and with drug-efficacy
in IBD pediatric patients. PACSIN2 is involved in vesicular trafficking and may affect the
release and content of extracellular vesicles (EVs), which mediate cell communication
and whose cargo modifies phenotypes of target cells. This study evaluates mechanisms
associating PACSIN2 polymorphism with interindividual variability in efficacy of thiopurines,
by considering the role of PACSIN2 in sorting specific miRNA in EVs.
Effects of stable PACSIN2 knock-down (KD) were evaluated in intestinal LS180 cells.
MTT cytotoxicity assay was used to verify mercaptopurine-sensitivity. EVs, released by
LS180 KD and MOCK control cells were isolated by ultracentrifuge and characterized by
nanoparticle tracking analysis (NTA). EVs miRNA-sequencing was performed by Illumina
Hi-seq 2000. EVs may alter drug cytotoxicity, therefore LS180 MOCK and KD cells were
co-treated with mercaptopurine and EVs. Statistical analysis was performed using t-test
and ANOVA.
Mercaptopurine was more cytotoxicity in LS180 KD cells (IC50 MOCK 3.23; IC50 KD 2.18
μM). No differences were observed by NTA in release of EVs between MOCK and KD cells
(t-test, p = 0.13). PACSIN2 KD altered intracellular and EVs expression of 6 and 24 miRNAs
respectively. EVs released by reduced mercaptopurine cytotoxicity (about 10%) and Rac1
protein expression in KD cells (ANOVA, p < 0.001), probably because they transport
different miRNAs.
In conclusion, PACSIN2 KD increase mercaptopurine cytotoxicity, probably, by deregulation
of miRNA expression in cells and EVs. These results will be further investigated to better
explain the link between PACSIN2 and EVs, whose miRNAs could provide a new scenario
in personalizing thiopurine treatment.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
Isolation, Characterization, Genome Analysis and Host Resistance Development of Two Novel Lastavirus Phages Active against Pandrug-Resistant Klebsiella pneumoniae
Get PDFKlebsiella pneumoniae is a global health threat and bacteriophages are a potential solution in combating pandrug-resistant K. pneumoniae infections. Two lytic phages, LASTA and SJM3, active against several pandrug-resistant, nosocomial strains of K. pneumoniae were isolated and characterized. Their host range is narrow and latent period is particularly long; however, their lysogenic nature was refuted using both bioinformatic and experimental approaches. Genome sequence analysis clustered them with only two other phages into the new genus Lastavirus. Genomes of LASTA and SJM3 differ in only 13 base pairs, mainly located in tail fiber genes. Individual phages, as well as their cocktail, demonstrated significant bacterial reduction capacity in a time-dependent manner, yielding up to 4 log reduction against planktonic, and up to 2.59 log on biofilm-embedded, cells. Bacteria emerging from the contact with the phages developed resistance and achieved numbers comparable to the growth control after 24 h. The resistance to the phage seems to be of a transient nature and varies significantly between the two phages, as resistance to LASTA remained constant while resensitization to SJM3 was more prominent. Albeit with very few differences, SJM3 performed better than LASTA overall; however, more investigation is needed in order to consider them for therapeutic application.Related to supp. material: [https://imagine.imgge.bg.ac.rs/handle/123456789/2072
Can we use biobanks to study infectious diseases?
Get PDFUnderstanding the molecular and environmental basis of diseases in order to improve
diagnosis and treatment represent a top priority for researchers. Much of the progress
occurred following the growth of various omics technologies and the IT progress
in developing large electronic databases capable of storing huge amounts of data.
Biobanks represents the most valuable resource for personalized medicine as these
are the large collection of various patient samples with well-annotated clinical data
which strive to identify possible links between genetic predisposition and disease. A
significant step forward are biobanks that are linked to the electronic health records of
each participant enabling up-to-date source of relevant medical information and those
“deeply phenotyped” for various other omics data, such as microbiome, epigenome,
transcriptome, metabolome and proteome.
Since infectious diseases still represent a huge threat to global human health, and host
genetic factors have been implied as determining risk factors for observed variations in
disease susceptibility, severity, and outcome, during this lecture we will discuss challenges
and opportunities of using biobanks as a potential source to study infectious diseases
based on the case example of isolated population-based longitudinal biobank “10,001
Dalmatians”. Results of a genome-wide association meta-analyses of 14 different
infectious-related phenotypes identified 29 infection-related genetic associations, most
belonging to rare variants, all of which have a role in immune response. These findings
support the concept that host genetic susceptibility to bacterial and viral infections in
adults is polygenic, where common variations have very low explained variance and/or
“unfortunate” combinations of numerous rare variants. Expanding our understanding
of rare variants may help in the construction of genetic panels which might predict an
individual’s lifetime vulnerability to major infectious diseases. Furthermore, longitudinal
biobanks are a valuable source of data for discovering host genetic variations involved
in infectious disease susceptibility and severity. Because infectious diseases continue to
exert selective pressure on our genomes, a global network of biobanks with access to
genetic and environmental data is required to further explain complicated mechanisms
underlying host-pathogen interactions and infectious disease vulnerability.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
A Similarity-based Normative Framework for Bio-plausible Neural Nets
Get PDFIn the last decade, Artificial Neural Nets (ANNs), rebranded as Deep Learning, have
revolutionized the field of Artificial Intelligence. While these neural nets have their origin
in analogy with the neural networks in the brain, in many ways they are trained in ways
that are very different from how real neurons learn. For example, to date there is no
satisfactory biologically plausible mechanism for backpropagation, the workhorse for
training ANNs.
Motivated by this gap, we have looked at alternative normative approaches to neural
networks that could give rise to more plausible learning rules. One such approach, which
works rather well for representation learning problems, is based on similarity matching
or kernel alignment. In this approach, one demands that similar sensory inputs produce
similar neural activities. From this rather limited constraint, one can give rise to interesting
neural networks performing many common unsupervised learning tasks. I will illustrate,
in particular, the case of representing continuous manifolds like spatial information. Here
, this approach produces representations very much like place cells in the hippocampus.
Consequences of our theory and its relations to some experiments would be discussed.
Time permitting, I would touch upon the role of similarity matching in current work in
ANNs as well.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
Clustering and classification of SARS-COV-2 isolates using RSCU
Get PDFThe existence of a large number of sequenced SARS-COV-2 isolates provides an
opportunity to observe genomic variability in a massive sample. The goal of our research
was to use data mining techniques to study possible correlation between codon usage
and classification by WHO-labels in a certain period of time.
The material includes 745,533 isolates with 12,236,672 coding sequences (proteins) from
NCBI (10.08.2022.). RSCU was used as a measure of codon usage. Samples are associated
with WHO-labels (based on Pango_Id) and time intervals. Inconsistency of WHO-labels
with periods in which the respective strains were actually present was observed. The
isolates with the observed discrepancy were excluded from the sample. Isolates without
assigned WHO-labels were also excluded. In addition, individual coding sequences
containing ambiguous nucleotide codes were eliminated.
Clustering was performed for each of the 12 common types of coding sequences
(proteins), with multiple methods and a different number of clusters. Neural clustering
gave the best results. For different protein types, different degrees of RSCU variability
are observed. In the case of proteins with a small variation in nucleotide contents, over
95% of the material belongs to a single cluster, while the other clusters are of negligible
size. In the case of proteins with more variations, a higher number of pure clusters (by
WHO-labels) is obtained, with a small number of heterogeneous clusters (about 10% of
the material). In those heterogeneous clusters, there are isolates with different WHOlabels
that were present in parallel at some point, as a kind of transitional forms between
two strains.
Different classification models were created on the same sample. Models based on
protein types with higher diversity between coding sequences are highly accurate (96-
100%). Using the classification models, the corresponding WHO-labels were associated
with isolates without previously assigned WHO-labels.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
From protein-protein to isoform-isoform interactions: the toolkit to map alternative splicing to interactome
Get PDFAlternative splicing (AS) can impact protein structure and lead to protein-protein
interaction (PPI) rewiring. Available PPI networks neglect alternative splicing isoforms: as
interactions might happen only between a subset of isoforms, the PPI network contains
both false-positive and false-negative interactions. Since it is not feasible to validate all
isoform-isoform interactions experimentally, we present a set of tools to investigate AS
on a network level: DIGGER to map splicing to the PPI network, as well as NEASE and
Spycone to evaluate the functional consequences of network rewiring.
DIGGER (https://exbio.wzw.tum.de/digger) integrates PPIs, domain-domain, and residuelevel
interactions - the structures that might be spliced in or out and result in interaction
gain or loss. Users can explore possible rewiring for an isoform or exon of interest and
extract relevant subnetworks. NEASE (https://github.com/louadi/NEASE) identifies pathways
that are significantly affected by network rewiring. NEASE extends classic gene set
enrichment analysis by considering isoform-specific interactions affecting pathways.
Spycone (https://github.com/yollct/spycone) addresses the time-course changes in AS. It
searches for isoforms that demonstrate similar temporal splicing patterns and reflect the
splicing co-regulation. Spycone further integrates gene set, network, and splicing-aware
NEASE enrichment.
Overall, we offer a splicing-focused network analysis toolkit that allows for studying the
mechanistic consequences of AS.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
Fatty Acid Data Analysis Unravels Skeletal Site and Age-Specific Features of Human Bone Marrow Adiposity
Get PDFAs adipose tissue (AT) undergoes metabolic reprogramming with age, we investigated
skeletal site-specific and age-dependent lipid profile of bone marrow adipose tissue (BMAT).
Acetabular and femoral BMAT, and gluteofemoral subcutaneous adipose tissue (gfSAT)
were obtained from matched osteoarthritis patients. Patients were classified into two
groups: younger (≤ 60 years) and aged (>60 years) adults. BMAT and gfSAT were explored
by using thin layer/gas chromatography coupled with cellular and molecular assays.
Data were interpreted and visualized by applying linear discriminant analysis (LDA) and
hierarchical clustering of fatty acid (FA) composition. Statistics was estimated by nonparametric
tests and Spearman’s rank correlation.
Analyses of total lipids revealed significantly reduced triglyceride content in femoral
(fBMAT) than in acetabular BMAT (aBMAT) and gfSAT. Frequencies of spontaneously
released saturated palmitic (C16:0) and stearic acids (C18:0) were higher in fBMAT than
in aBMAT and gfSAT (p=0.036 and p=0.046, n=8). Cluster heatmap and LDA showed that
fBMAT differed to acetabular and gfSAT, while acetabular and gfSAT were more similar in
FA profiles. FA profiles of AT depots varied with patient’s age. Contribution of palmitic acid
was increased in aged group in all AT depots, while stearic acid declined in aged group in
BMAT compartments only. fBMAT cellularity declined with age (r=-0.675, n=14, p=0.037).
Additionally, the presence of CD45-CD31-CD34+CD24+ adipogenic progenitor (stem) cells
was increased in fBMAT (0.46±0.03%) when compared to aBMAT (0.21±0.01%) depot.
Femoral mesenchymal stem cells displayed pronounced adipogenesis comparing to their
acetabular counterparts.
Our findings suggest that specific lipid profile of fBMAT imposes adipogenic commitment
of stem cells within this skeletal site.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202
Comparative study of in silico protein design techniques
Get PDFProtein design plays a pivotal role in various scientific and industrial applications, such
as drug development and biotechnology. With the advancement of computational
methods, new tools and algorithms have emerged to facilitate the generation of novel
protein designs. This study presents a comparative analysis of Pepspec and RFdiffusion,
two prominent methods in protein design, to evaluate their effectiveness in designing
peptides with desired properties. Mainly, we aim to design peptides that bind with high
affinity and specificity to a desired protein target.
Pepspec is an application native to the Rosetta software package. It relies on Monte Carlo
sampling of backbone confirmations and residue mutations and a stochastic optimization
based on the Rosetta score – a measure approximating the binding free-energy of the
complex.
On the other hand, a recently developed tool, RFdiffusion, is a denoising diffusion
probabilistic model based on an existing artificial neural network, RoseTTAFold, developed
for protein structure estimation. It is trained to remove noise from protein structures
on a large database of protein complexes to ultimately be able to generate novel binder
designs based on the target structure.
In this study, we aim to compare the efficiency of these two design tools. As it is common
in generative ML algorithms, the comparison will be made by evaluating both the design
quality and design versatility. The quality will be assessed by using the well-known
AlphaFold2 Machine learning tool to estimate the binding affinity of the peptide-protein
complex while the versatility will be measured using standard sequence based statistical
methods.
RFdiffusion and Pepspec offer distinct approaches to protein design. By assessing
the strengths and limitations of each method in this study, we aim to deepen the
understanding of these methods and allow leveraging these tools effectively in designing
peptides with desired characteristics, contributing to advancements in the field of protein
engineering and biotechnology.Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 202