Score Coefficients Optimization

Score coefficients of 1 can be used by default to rank candidate molecular formulas. However, to achieve more desirable formula assignments, identification ranking score coefficients (PARAM0023) should be computed using a set of positive weights (see section S.1). To optimize these weights, a set of highest quality annotations from authentic standards should be fed in the IDSL.UFA workflow using the score_function_optimization tab in the UFA parameter spreadsheet. PARAM0007 in the parameters tab should be also selected YES to run this tab.

Note: The IDSL.UFA pipeline originally was developed to annotate IDSL.IPA peaklists that were generated using ¹²C/¹³C isotopologue pairs by the IDSL.IPA pipeline. Nevertheless, IDSL.UFA can still annotate IDSL.IPA peaklists with non-carbon ion pairs when the fifth coefficient in the score function in the equation 6 in the main manuscript is zero to neutralize interferences of R¹³C values. The score coefficients can be adjusted through PARAM0023 in the UFA parameter spreadsheet.

Score Coefficients Corrector

A UFA_score_coefficients_corrector function is included in the IDSL.UFA package to re-evaluate the ranks of annotated formulas using updated score coefficients. By updating the coefficients, this function can provide a refined ranking of the formulas to prioritize the most relevant or accurate formulas using optimized score coefficients.

UFA_score_coefficients_corrector(input_annotated_molf_address, output_annotated_molf_address,
				 scoreCoefficients, number_processing_threads = 1)

input_annotated_molf_address: Address to load the individual Molecular Formula Annotation Tables

output_annotated_molf_address: Address to save the individual Molecular Formula Annotation Tables

scoreCoefficients: A vector of five numbers representing coefficients of the identification score function

number_processing_threads: Number of processing threads for multi-threaded computations