Bioinformatics Workflow for Wet-lab Molecular Biologists

A wet-lab bioinformatics workflow is a short, repeatable path from a sequence file to a cloning or PCR decision: annotate the map, compare it with reads or references, and store the evidence with the experiment. It is not a substitute for a genome-center NGS pipeline.

This article is for molecular biologists who live in plasmids, primers, and Sanger traces and need a lighter informatics habit that still survives review.

What Wet-Lab Bioinformatics Actually Has to Produce

At the bench, informatics succeeds when it answers four questions. What molecule did we intend? What molecule do we have? Where do they differ? Can someone else reconstruct that comparison in six months?

Those questions map to files, not to a vague "I BLASTED it." You need an annotated reference (usually a plasmid map), one or more reads or assemblies, an alignment with visible mismatches, and a written decision: ship, rebuild, or resequence.

Start From an Annotated Map, Not From a Naked FASTA

FASTA is a sequence. A cloning decision needs features: promoter, ORF, spacer, primer binding sites, and expected junctions. If the only file in the folder is "gene.fasta", you cannot tell whether a mismatch sits in a linker or in an active site.

Import or build the plasmid in sequence software, then freeze a version name (pABC-v3) before oligos go out. Every later alignment should use that frozen file, not an edited copy someone "cleaned up" after the fact.

Shared libraries help when several people clone from the same backbone. A plasmid library is a starting catalog; the project still needs its own versioned working copy once you modify a cassette.

Align Reads in the Same Frame as the Experiment

Sanger verification is still the default for cloning junctions. Align traces to the expected plasmid, not to a genome, unless you are checking a genomic edit. Inspect mixed peaks at the spacer or insert boundary; a "99% identity" number can hide a mixed colony.

If you use whole-plasmid sequencing, treat the assembly as a hypothesis. Confirm that the assembler did not drop a repeat in the origin or in a duplicated terminator. Spot-check the junctions you actually designed, especially Gibson overlaps and Type IIS scars.

BLAST is a search tool. Use it to identify an unknown insert or to see whether a primer hits something obvious. Do not use a BLAST hit as clone QC in place of an alignment to your own reference.

Name Files So the Notebook Can Find Them

A workable convention is project, construct, version, role, date: PRJ17_pET-Kinase_v3_Sanger-F_2026-08-12.ab1. Role might be design, oligo, sanger, or assembly. Put the same string in the ELN entry.

Do not store the only copy in a personal Downloads folder. Cloud project storage or an ELN attachment with permissions is the minimum. Zettalab is one example of a workspace that can keep the map, the alignment, and the experiment record in the same project so the pass/fail note is not a chat message.

Write the Decision, Not Only the File Path

The bioinformatics step is unfinished until a human writes: matches v3 at both junctions; silent T>C in codon 14, accepted; or mixed peaks at the spacer, colony discarded. That sentence is what a PI or a future you will read.

If the construct is rejected, keep the traces anyway. Failed clones teach primer and assembly problems. Deleting them because they are "messy" removes the only evidence of a systematic error.

FAQ

Do wet-lab scientists need a bioinformatics workflow if they do not run NGS?

Yes. Cloning, PCR, CRISPR plasmid builds, and Sanger QC are already informatics. The workflow is smaller than an RNA-seq pipeline, but it still needs a reference, an alignment, and a stored decision. Labs that skip this end up with freezer boxes of plasmids that do not match maps. You do not need to become a computational biologist. You need a repeatable file habit and software that can show mismatches on an annotated construct. Write the pass/fail sentence in the same record as the files so the decision cannot drift from the traces.

What is the minimum software stack for this workflow?

You need a map editor that understands features, an aligner that accepts Sanger or FASTA reads, and a place to store files with names and permissions. That can be one connected platform or three tools with a strict folder convention. Evaluate whether primer features and translations are visible during alignment. A viewer that only shows colored nucleotides without annotation will miss junction errors that sit in "correct-looking" DNA. If three tools are used, write which file is the system of record so people do not edit the copy.

How should Sanger mixed peaks be handled?

Treat mixed peaks at a designed junction as a failed clone until you re-streak and resequence a clean colony. Mixed peaks in a homopolymer or at the end of a weak trace may be chemistry, not biology; resequence with a closer primer before rebuilding. Record both the chromatogram and the interpretation. Do not clean mixed bases to the expected nucleotide in a "consensus" FASTA that you then treat as truth. If you re-streak, give the new colony a new ID rather than overwriting the failed trace.

Where should alignment files live relative to the ELN?

They should be one click from the experiment that produced the colony. Attachments on the record, or links to a project folder with the same ID, both work. What fails is a Slack screenshot with no raw ab1. If the lab uses a cloud ELN, set permissions so collaborators can read traces without editing the reference map. Connected R&D workspaces reduce that split; the requirement is reconstructability, not a particular vendor. Keep the raw ab1 even after you export a pretty PDF alignment.

How is this different from a core-facility bioinformatics pipeline?

Core pipelines ingest libraries, run batch QC, and emit genome-scale results. Wet-lab construct QC ingests one map and a handful of reads and emits a clone decision. Using an NGS pipeline UI to "align a plasmid" often hides features you care about, such as a 4-bp Golden Gate scar. Keep genome work in genome tools. Keep plasmid work in plasmid tools. Hand off only when a result truly needs genomic context, such as confirming a knock-in on-target. The clone decision still belongs in the molecular biology notebook, not only in a bioinformatics ticket.

Conclusion

Wet-lab bioinformatics is a cloning quality system: annotated reference, raw reads, alignment, and a written pass/fail. Keep those four objects together, named, and permissioned. Platforms that connect maps with notebook entries, including Zettalab, make that bundle easier to keep intact. For day-to-day sequence handling, use the Zettalab sequence and ELN guide as a starting checklist, then freeze your own construct versions before the next oligo order.