When a label looks like a calendar date
A spreadsheet does not always treat incoming text as text. To save users from typing dates in a rigid format, it tries to recognize familiar patterns automatically. A scientific identifier that resembles a month-and-number combination can therefore be interpreted as a date when it is pasted, opened from a file, or entered into a cell.
That behavior has affected gene symbols, including names such as MARCH1 and SEPT2. To a researcher, each is a fixed label. To the spreadsheet’s parser, it can look like a date worth converting into the program’s internal date value. The cell may then display a calendar date—or a number—rather than the original identifier.
Why the conversion can damage a dataset
Once the spreadsheet has replaced the label with a date value, changing the cell’s display format may not bring the original characters back. The program has not merely changed how the text looks; it may have discarded the text and stored a different kind of value. If the altered file is later used for analysis or shared with collaborators, the changed identifier can be mistaken for a valid label, disrupting matching and comparisons.
The risk is easy to miss because the rest of the sheet can look normal. A single converted entry among thousands may survive unnoticed through several rounds of editing and reuse.
Keeping identifiers intact
The safest approach is to tell the spreadsheet what the data means before it interprets it: import the relevant column explicitly as text, or set its cells to a text format before entering the values. Simply saving a file as CSV does not reliably protect identifiers; CSV stores characters without instructions about their intended type, and spreadsheet software may guess again when reopening it.
Some researchers have renamed affected items as a workaround. But preserving the original identifier as text avoids creating a second naming system—and prevents a convenient guess by the software from quietly becoming a scientific data error.
