Do the 5'untranslated domains of human cDNAs challenge the rules for initiation of translation (or is it vice versa)?

The validity of the scanning mechanism for initiation of translation has been questioned based on a compilation of human cDNA sequences that showed a high frequency of upstream ATG codons. However, closer scrutiny of those cDNAs upholds the opposite view: the 5'UTRs on most cDNAs are compatible with standard rules for initiation of translation, and those rules can be used to flag anomalous cDNAs that, upon checking, turn out to have been misinterpreted. Some of the problematic 5'UTR sequences that persist, after obvious errors in the cDNA library have been corrected, might derive from transcripts that are not intended to be translated. Examples are given of genes that, for regulatory reasons, produce transcripts that are truncated, or retain an intron, or are otherwise configured in a way that precludes translation. The existence of a cDNA proves that a gene is transcribed, but only that; not every cDNA derives from a functional mRNA. Along with providing practical guidelines for interpreting cDNA sequences, the scanning model provides a theoretical framework for understanding the effects of certain mutations in the 5'UTR that alter the translatability of mRNAs, thereby contributing to cancer and other human diseases.