https://github.com/microsoftgenomics

Address of the bookmark: https://github.com/microsoftgenomics

More at http://www.nature.com/articles/srep24175

Address of the bookmark: https://sourceforge.net/projects/sb2nhri/files/MyCC/

https://academic.oup.com/bioinformatics/article/32/4/605/1744462/MaxBin-2-0-an-automated-binning-algorithm-to

The most recent version of MaxBin is 2.2, which supports the analysis of coassemblies of multiple samples. It is available at this JBEI downloads sites as well as MaxBin and MaxBin 2.0 sourceforge sites.

Address of the bookmark: http://downloads.jbei.org/data/microbial_communities/MaxBin/MaxBin.html

GroopM is largely parameter-free. Use: groopm -h for more info.

For installation and usage instructions see : http://ecogenomics.github.io/GroopM/

Address of the bookmark: https://github.com/ecogenomics/GroopM

JSON is built on two structures:

• A collection of name/value pairs. In various languages, this is realized as an object, record, struct, dictionary, hash table, keyed list, or associative array.
• An ordered list of values. In most languages, this is realized as an array, vector, list, or sequence.

These are universal data structures. Virtually all modern programming languages support them in one form or another. It makes sense that a data format that is interchangeable with programming languages also be based on these structures.

Address of the bookmark: http://json.org/

Address of the bookmark: https://chunlab.wordpress.com/tetra-nucleotide-analysis/

Dr Altan Kara is a Bioinformatics specialist at the faculty of Gene Engineering and Biotechnology Institute at TUBITAK MAM Research Center. His research interest revolves around the cancer informatics and computational aided-drug design. I applaud Dr Altan for clearly setting out both his expectations of people that join his lab/university in addition to listing his responsibilities to his research members at TUBITAK MAM Research Institüte. Hopefully, this interview will prove useful to others in the field, especially to those who are just starting their bioinformatics careers.

You can find out more about Dr Altan by visiting his (well documented) lab page (http://gmbe.mam.tubitak.gov.tr/en) and BOL page http://bioinformaticsonline.com/profile/altan . And now, on to the BOL:“Tryst with a Bioinformatician” interview series ...

• What push you to join Computational Biology/Bioinformatics?

According to me, bioinformatics is the center of modern biological research and if a researcher wants to discover new biological insights by evaluating the globally produced biological data to derivate unified solutions for specific biological problems, learning bioinformatics is the only way to achieve this goal.

• What fascinates you about Computational Biology/Bioinformatics?

It's flexibility. As well known, there are highly diverse and complex biological questions are waiting to be enlightened and it's impossible to bring solutions to this diversity by using similar approaches. Thus, the employed method has to be unique for the targeted biological problem and by using bioinformatics tools this can be easily achieved. 

• What is the one word you would use to describe yourself?

Bioinformatician. :)

• Can you please describe your research work in a nutshell for BOL users.

At my current Institute, I am working in the field of cancer bioinformatics. Briefly, the overall aim of the project which I am working for (AKMARK (Project CODE:5153403)) is, applying a bioinformatics-supported genome, transcriptome, proteome, and metabolome analysis to reveal the molecular profile of the disease through an integrated approach, and to develop an early diagnosis and scanning kit based on this profile. Alterations in the gene, transcript, protein, and metabolite profiles between normal tissue, normal tissue adjoined to the tumor (reactive stroma), tumor tissue, lymph node metastasis, and blood samples taken from the same patient and the reflection of these changes in some other selected body fluids will be revealed within the scope of the project. The molecular structures involved in the development and progression of NSCLC will be determined and relations with the clinical, tumor-node-metastasis (TNM) staging and histology will be made. The development of a diagnostic kit for immediate clinical purposes and an electrochemical biosensor for quick on-site applications are targeted through the development of a number of antibody and aptamer formed against the most specific biomarker selected from the panel.

• Is there anything else we should know about you and your research?

Besides AKMARK, I am also in preparation of having a side project that aims for the development of a computational method to design inhibitors for prokaryotic two-component systems. In this project, I will be in collaboration with Prof. Maria Kontoyianni, SIUE: Southern Illinois University Edwardsville, School of Pharmacy.

• What was your greatest scientific disappointment in life till now?

So far I do not experience any memorable scientific disappointment in my life. :)

• What major research challenges and problems did you face yet? How did you handle them?

The major challenge which I faced so far in my scientific career was predicting the interaction between the prokaryotic two-component proteins. To be able to accurately predict the interactions between these proteins, I create a meta-predictor by using a support vector machine. By using this technique I integrated six different protein-protein interaction methods in a way to cover disadvantage of one method with the advantage of another one. The meta-predictor which I developed during this work is accessible via http://metapred2cs.ibers.aber.ac.uk/ and for more detailed information about the system the articles with the PMID IDs; PMID: 27378293 and PMID: 26384938 can be read.

• What's your all-time favourite bioinformatics package, and why?

For me, the best bioinformatics package is R/Bioconductor. The reason why I like this package is, it provides lots of useful tools for comprehensive analysis and comparison of high-throughput experimental data in an integrated manner and besides lots of the packages it provides, it is open source and also open for development. As a result, it provides strong and flexible ways to do science.

• In bioinformatics, do you see yourself in which of the following roles-scientist, analyst, developer, engineer or pure academician?

Scientist / Developer.

• What will you like to accomplish in next five years / ten years?

For my current research, I would like to design a pipeline to automatically integrate and analyse omics data for cancer research which will be specifically aiming for biomarker and novel drug target discovery. In addition to this, I also like to develop another pipeline for prokaryotic TCS protein structure prediction and inhibitor design.

• When you will be retired, what would you tell next generation bioinformaticians?

Bioinformatics is not all about scripting and researchers who study in this field should never expect a tool to do their analyses for them. Besides computational skills, a bioinformatician must have a strong biological background in his/her research area which will allow them to understand if anything went wrong during their run by only looking at the results instead of just blindly trusting the output of the bioinformatics tools.

• What you always miss in bioinformatics when you will no longer working in this field?

Bioinformatics is open to doing multi-discipliner research with scientists all around the world. As a result, while I studying in this field I can interactively learn a lot from wide range research community. I think this is the one thing which I will miss the most.

• If there will be bioinformatics company owned by you in future, What are your company focus and aim?

With the increasing amount of data in databases, there is already a massive need for effective methods to eliminate the manipulated data and reach to clean/useful information. As days pass, the requirement of data mining will be the first step of any research project. For this reason, the major goal of my bioinformatics company will be developing effective tools to eliminate manipulated datasets and information that exist in the literature and provide trustworthy clean information/datasets for researchers.

• How much bioinformatics change in 2050, according to your wild imagination?

Bioinformatics is a field that constantly and dynamically changes. As the bioinformatics progress, new tools and methods become available and they provide a better application of existing methods or totally new methods that offer an alternative solution to various biological problems. A long with these updates, developers also provide easy to use GUIs for most of the tools. Considering this, if the field carries on developing like this, every single researcher with a strong biological background can be able to perform bioinformatics analyses by him/herself without needing a professional help. As a result, almost all of the bioinformaticians will be responsible just for development of new methods/tools.

• What would one piece of advice you give someone who's trying to reinvent themselves and enter into bioinformatics sector?

Bioinformatics is a wide field with a lot of career options. Thus, if a researcher likes to step into this field first he/she should be clear about the branch of the bioinformatics they like to study in. Following to this decision they should first learn at least one programing language and investigate the ways of how other researcher employed that language in their researches and WHY? A researcher, in this field, should never create and use copy paste scripts but always must understand WHY the other researcher worked in that way. Knowing the answer of this question is the only way to learn bioinformatics. Besides, a researcher in the field of bioinformatics (from any branch) must always be good about the environmental control. In other words, one should always easily control input output directories, modify files or directories, annotate and modify employed scripts during the research and should not allow any confusion during the different stages of the research. Finally, they should not blindly trust the output of a tool/software but do a benchmarking test for each of the tools which they decided to utilise in their research. In addition to this, even if the tools pass the benchmarking, researchers should have a good biological background in their field to tell if anything when wrong during the process by only looking the output(s) of the employed pipelines/packages/tools.  

Genome annotation is a multi-level process that includes prediction of protein-coding genes, as well as other functional genome units such as structural RNAs, tRNAs, small RNAs, pseudogenes, control regions, direct and inverted repeats, insertion sequences, transposons and other mobile elements.

NCBI has developed an automatic prokaryotic genome annotation pipeline that combines ab initio gene prediction algorithms with homology based methods. The first version of NCBI Prokaryotic Genome Automatic Annotation Pipeline (PGAAP; see Pubmed Article) developed in 2005 has been replaced with an upgraded version that is capable of processing a larger data volume. You can find a more detailed description of the new version of the pipeline in NCBI Handbook chapter. NCBI's annotation pipeline depends on several internal databases and is not currently available for download or use outside of the NCBI environment.

https://www.ncbi.nlm.nih.gov/genome/annotation_prok/

Address of the bookmark: https://www.ncbi.nlm.nih.gov/genome/annotation_prok/

https://bio7.org/about/

Address of the bookmark: https://bio7.org/home-2/

1. Heatmaps: Exploring Patterns in High-Dimensional Data

Heatmaps are a go-to visualization for representing high-dimensional datasets, such as gene expression or metabolomics data. They use color gradients to display data intensity, making patterns and clusters easily detectable.

• Applications: Gene expression analysis, pathway enrichment, methylation studies.

• Tools: Seaborn (Python), ComplexHeatmap (R), Morpheus (web-based).

Tip: Add dendrograms to visualize clustering of rows and columns for hierarchical relationships.

2. Volcano Plots: Highlighting Differential Features

Volcano plots are indispensable for identifying significantly differentially expressed genes or proteins. They plot the log2 fold change against –log10(p-value), making it easy to spot statistically significant changes.

• Applications: RNA-seq, proteomics, and metabolomics.

• Tools: ggplot2 (R), EnhancedVolcano (R), Plotly (Python).

Tip: Use color to highlight significant features and label key genes or proteins.

3. PCA Plots: Reducing Complexity with Principal Component Analysis

Principal Component Analysis (PCA) plots are used to reduce dimensionality and uncover trends or clusters in data. They provide insights into sample variability and grouping.

• Applications: Transcriptomics, metabolomics, microbiome studies.

• Tools: scikit-learn + Matplotlib (Python), prcomp (R), ClustVis (web-based).

Tip: Annotate clusters with metadata to enhance interpretability.

4. Manhattan Plots: Genome-Wide Association Studies

Manhattan plots visualize p-values across the genome, making it easy to identify significant associations in genome-wide studies. They resemble city skylines, with the highest peaks indicating loci of interest.

• Applications: GWAS, QTL mapping.

• Tools: qqman (R), Matplotlib (Python).

Tip: Use alternating colors for chromosomes and highlight significant SNPs for clarity.

5. Circular Plots (Circos): Visualizing Genomic Relationships

Circular plots are ideal for visualizing relationships across the genome, such as structural variations, gene duplications, or synteny.

• Applications: Comparative genomics, structural variation studies.

• Tools: Circos (standalone), Rcircos (R), pyCircos (Python).

Tip: Keep the plot clean and avoid overcrowding to maintain readability.

6. Sankey Diagrams: Tracking Data Flows

Sankey diagrams visualize flows or relationships between categories, often used to track changes in gene expression or pathway enrichment across conditions.

• Applications: Pathway analysis, gene set enrichment analysis.

• Tools: Plotly (Python), networkD3 (R).

Tip: Use gradients or distinct colors to highlight key transitions.

7. Network Graphs: Mapping Interactions

Network graphs represent relationships between entities, such as protein-protein interactions or gene regulatory networks. Nodes represent entities, and edges represent relationships.

• Applications: Systems biology, interactomics.

• Tools: Cytoscape (standalone), igraph (R), NetworkX (Python).

Tip: Use edge thickness or node size to represent interaction strength or centrality.

8. Violin Plots: Visualizing Data Distribution

Violin plots combine a boxplot with a density plot, showing the distribution and variability of data.

• Applications: Single-cell RNA-seq, quantitative trait analysis.

• Tools: Seaborn (Python), ggplot2 (R).

Tip: Split violins by groups for side-by-side comparisons.

9. Time-Series Plots: Monitoring Changes Over Time

Time-series plots display changes in variables across time points, useful for tracking gene expression dynamics or metabolic fluxes.

• Applications: Time-course experiments, cell cycle studies.

• Tools: Matplotlib (Python), ggplot2 (R).

Tip: Smooth the data to highlight trends while avoiding overfitting.

10. Genome Tracks: Visualizing Genomic Features

Genome tracks display multiple layers of genomic data, such as gene annotations, sequencing coverage, and epigenetic marks.

• Applications: ChIP-seq, ATAC-seq, whole-genome sequencing.

• Tools: IGV (standalone), pyGenomeTracks (Python).

Tip: Stack related tracks for direct comparisons.

11. UpSet Plots: Visualizing Set Intersections

UpSet plots are a powerful alternative to Venn diagrams for visualizing intersections between multiple datasets.

• Applications: Overlap analysis for gene sets, pathways, or variants.

• Tools: UpSetR (R), ComplexUpset (Python).

Tip: Use bar plots to represent the size of each intersection for added clarity.

12. Ridge Plots: Comparing Distributions

Ridge plots visualize the distributions of multiple datasets, stacked for easy comparison.

• Applications: Transcriptomics, single-cell RNA-seq.

• Tools: ggridges (R), Matplotlib (Python).

Tip: Use transparency and consistent scaling for better readability.

13. Chord Diagrams: Visualizing Connections Between Groups

Chord diagrams illustrate relationships between categories, such as shared genes between pathways or overlaps in regulatory elements.

• Applications: Pathway overlap, synteny, co-expression networks.

• Tools: Circlize (R), Holoviews (Python).

Tip: Use distinct colors for each group to emphasize relationships.

14. Treemaps: Hierarchical Data Representation

Treemaps visualize hierarchical data as nested rectangles, with area proportional to data size.

• Applications: Ontology enrichment, pathway analysis.

• Tools: Treemapify (R), Plotly (Python).

Tip: Use colors to represent additional variables, like significance or enrichment scores.

15. T-SNE/UMAP Plots: Dimensionality Reduction for Clustering

T-SNE and UMAP plots are great for visualizing high-dimensional data in two dimensions while preserving local or global structure.

• Applications: Single-cell transcriptomics, clustering analyses.

• Tools: scikit-learn (Python), Seurat (R).

Tip: Combine with metadata annotations for better cluster interpretation.

Bringing It All Together

The choice of visualization can significantly impact the insights gained from bioinformatics data. By selecting plots tailored to your data type and analysis goals, you can effectively communicate your findings and make your research more impactful. Whether you’re a seasoned bioinformatician or a beginner, mastering these visualizations will elevate your analyses and presentations.