| Data type | What it captures | Planning use |
|---|---|---|
| Genome sequence | The complete set of genetic information of an organism | Choose it when the question concerns an organism’s whole genetic content. Sequence comparison tools, covered below, are the natural next step. |
| Transcriptome (RNA-seq) | The complete set of RNA molecules expressed in a cell or tissue at a specific time | Plan the sampling time carefully, because the set describes one moment. RNA-seq analysis can then look for differentially expressed genes and splicing. |
| Protein structure | Structure data from X-ray crystallography and similar methods | Use it when the question is about molecular shape, not only sequence. |
| Mass spectrometry spectra | Spectra used to identify proteins and metabolites | Use it for identifying the proteins, the “worker molecules” of a cell, and the metabolites in a sample. |
| Imaging data | Images from MRI, CT and microscopy | Include it when the evidence is image-based, and expect an analysis route separate from sequence work. |
Definitions follow LMU’s bioinformatics overview, and the RNA-seq analysis point comes from GeeksforGeeks’ summary of bioinformatics applications.
What is bioinformatics? It is a scientific field that uses computers, mathematics and statistics to collect, store, organize and analyze biological, medical and health information. The data type you begin with decides which tool you need.
This guide follows one practical path: pick a data type, match it to a tool, read the output with care, then build skills in steps. Study-design questions such as how do clinical trials work sit in the research methods hub, so this page stays with data and software.
What bioinformatics changes in practice
Bioinformatics changes who can answer a biological question. LMU describes it as a bridge science between biology, medicine and computer science, so a laboratory measurement is the start of an analysis rather than the end of a project.
Pick the framing below that matches your goal, because each one points to different skills and data.
- If your interest is medicine: analyzing individual genetic profiles can aid the development of personalized treatment plans, according to a Patsnap Synapse overview. This suits readers drawn to patient-linked data and genetic variation.
- If your interest is traits and variation: genome-wide association studies use bioinformatics approaches to find genetic variation linked to traits, as GeeksforGeeks notes. This suits readers who like working with large comparisons across genomes.
- If your interest is gene activity: expression data show which RNA is present, and RNA-seq analysis tools look for differentially expressed genes and splicing. The regulation behind that activity is explained in What Is Epigenetics? How Genes Are Switched On and Off — A Clear Guide For 2026 – Biolifetechnologies.
- If your interest is the laboratory side: teams work best when the analyst understands where the data came from. LMU states that a bioinformatician needs the biological background of the data to choose suitable algorithms.
The practical consequence is that the choice of career path, project and tool all begin with the biology of the data. For a neighbouring research topic, read Human Brain Organoids: What They Are, How They’re Made, and Why They Matter In 2026 – Biolifetechnologies. The next section turns to the tools themselves.
Matching a data source to a tool
Sequencing and other instruments produce raw data, and software turns it into something a researcher can interpret. For sequence data, the best-known entry point is BLAST, the Basic Local Alignment Search Tool from NCBI. BLAST finds regions of similarity between sequences. It compares nucleotide or protein sequences and calculates the statistical significance of matches, according to a UC Berkeley library guide to NCBI BLAST.
The four NCBI WebBLAST search types
NCBI’s WebBLAST offers four main search types. Choose by what you hold in hand and what you want to compare it with.
- BLASTn: compares nucleotide query sequences with nucleotide sequences. Choose it when you hold DNA or RNA sequence and want matching sequence.
- BLASTx: translates a nucleotide query in six reading frames and searches a protein database. The Berkeley guide notes it is often the first analysis performed on a newly determined nucleotide sequence, so it suits a sequence whose function you do not yet know.
- BLASTp: compares protein queries with protein sequences. It is useful for identifying a protein you already have.
- tBLASTn: searches a protein query against six-frame translations of nucleotide databases. It is useful for finding homologous protein-coding regions in unannotated nucleotide sequences.
Choosing between them
Match the query type to your starting material, not to the tool you have heard of most. A nucleotide sequence of unknown function points to BLASTx. A known protein searched against raw, unannotated DNA points to tBLASTn. If query and target are the same kind of molecule, BLASTn or BLASTp is the direct route.
BLAST covers sequence questions only. Expression, spectra, structures and images each need their own analysis route, which is why the table at the top of this guide starts from the data type and not from the software.
Reading BLAST results and NCBI records
A BLAST hit is only useful if you can judge it. For significant alignments, the E value should be very close to zero, which means a larger E value is a reason for caution, not a result to report. The percentage similarity tells you how close two sequences are, while the significance statistic tells you whether the match is meaningful, so read both.
Record the accession number of every sequence you use. An accession number is a unique identifier assigned to records in the NCBI databases, and it lets you or a colleague return to exactly the same record later.
When you need a trustworthy comparison target, look for RefSeq records. These are high-quality sequences curated and annotated by NCBI staff.
The Berkeley guide gives a worked example from the human mitochondrial genome. Three RefSeq reference sequences exist for it: modern human, Neanderthal and Denisovan. A modern human query is reported as 99% similar to the Neanderthal sequence and 98% similar to the Denisovan sequence. The lesson for beginners is that a genome-scale question can be explored with a single, well-chosen reference and a standard search.
A beginner learning path tied to real data types
Start narrow. The aim of a first project is to understand one data type and one tool well enough to explain the output, not to cover the whole field.
- Learn the biology of one data type first. Because analysts must understand the biological background of the data to choose suitable algorithms, spend your first effort on what sequences, expression sets or spectra actually represent.
- Run a guided BLAST search. A LabXchange video tutorial walks through NCBI BLAST step by step, covering nucleotide and protein searches and how to interpret the results. It suits readers who prefer watching a search before running one.
- Practise on records, not only on searches. Open the matching NCBI record for a hit, note its accession number, and see whether a curated RefSeq version exists.
- Add a second data type once BLAST output makes sense. RNA-seq expression data are a logical next step, since bioinformatics tools analyze them for differentially expressed genes and splicing. Read the table above for how a transcriptome differs from a genome sequence.
Wait on structure and imaging work until you are comfortable with sequence records. They are separate analysis routes and add little to a first search.
Frequently asked questions
Why is software needed to analyze biological data?
LMU explains that the amount of biological data is usually far too large to be evaluated by hand. The field therefore develops methods and software tools, especially for large and complex data sets.
Can I trust results from any dataset?
Not automatically. Bioinformatics relies heavily on data generated by experimental techniques, and that data may vary in quality and accuracy, so the analysis can only be as dependable as its input.
How is mitochondrial DNA inherited?
It is inherited only through the maternal lineage and has a low rate of recombination, according to the UC Berkeley BLAST guide. That makes it a clear subject for sequence comparison.
