Heuristic

21 methods · 1981–2021

Heuristic: Hand-engineered interpretation rules rather than one global optimisation: extend a sequence peak by peak, enumerate the compositions that fit the mass, score what survives. The oldest approach here, and the closest to reading by hand.

Hand-engineered interpretation rules rather than one global optimisation: extend a sequence peak by peak, enumerate the compositions that fit the mass, score what survives. The oldest approach here, and the closest to reading by hand.

The earliest of its 21 methods is PAAS (1981); 20 more have followed.

Methods 21
Papers describing them 31
Authors 92
Active 1981-04-01 to 2021-11-23
Kinds algorithm (21)
Acquisition DDA (20)

Methods (21)

Oldest first, by the paper that describes each one.

  • PAAS (1981): The 1981 original, and the earliest de novo peptide sequencing program in this catalog: a computer program that estimates the possible amino acid sequences of a peptide from its mass spectrum, with no database to search. It could also read a sequence from the PTH amino acids released by Edman degradation, which is what sequencing looked like before tandem MS.
  • Improved PAAS (1983): The 1983 revision of PAAS, by the same Osaka group with Wollnik added. Sits between the 1981 original and PAAS 3, so the lineage spans three papers in four years.
  • PAAS 3 (1984): 1984 FORTRAN program that determines the probable amino acid sequence of a peptide from the sequence ion peaks of its mass spectrum, without a database. The third of the Osaka group’s lineage, after PAAS (1981) and Improved PAAS (1983), and the most cited of the three.
  • SEQPEP (1989): Computer program for interpreting high-energy collision tandem mass spectra of peptides.
  • Lutefisk (1997): First widely-used heuristic de novo tool
  • SeqMS (1998): Software aid for automated interpretation of low-energy CID spectra for de novo sequencing by tandem MS.
  • 2D Fragment Correlation (2000): 2D fragment correlation MS
  • PepSeq (2000): Interactive de novo peptide sequencing application in BioLynx, the biopolymer analysis component of the MassLynx suite. The manual is explicit that it is user-driven rather than automatic: PepSeq “has been designed as an interactive program that allows the user to make decisions at each step in deducing the sequence of a peptide”. It scores and annotates spectra on the a, b, y and z ions, and takes input either from a MassLynx data file or as a plain mass and intensity list. Three modes: typing a candidate sequence to see its theoretical fragments annotated against the spectrum; FindTag, which builds a set of sub-sequences whose series ions best match the spectrum and extends them step or leap-wise until the molecular weight matches the user-supplied precursor; and MassSeq, a separately purchased option that performs the sequencing automatically from the precursor mass, modifications and a mass-accuracy estimate. There is no methods paper: the method is documented only in the vendor manual, which is catalogued here as the accompanying resource. Provenance runs Micromass UK Ltd to Waters Corporation, and the Version 4.0 guide catches the handover in progress, carrying a Waters part number and a Micromass part number side by side over a 1993-2002 Micromass UK Ltd copyright.
  • Sequence pattern correlation (2002): Deduces peptide sequences without a database by cross-correlating an experimental CID spectrum with predicted sequence patterns from y and non-y fragmentation, claimed insensitive to mass tolerance.
  • MALDI/TOF-TOF de novo sequencing (2002): Sequence-extension approach for de novo sequencing from MALDI TOF-TOF spectra.
  • MALDI-QIT de novo sequencing (2003): Preliminary MALDI-QIT de novo
  • Sequit (2004): Software for de novo peptide sequencing by MALDI post-source decay mass spectrometry.
  • Multi-pass CID spectrum interpretation (2005): Interprets low-energy CID product ion spectra from an ion trap in several passes, extracting the data rigorously and then building the sequence, rather than scoring candidate paths in one sweep.
  • Proteomics-grade de novo sequencing (2005): Proteomics-grade de novo sequencing approach for large MS/MS data sets.
  • Adaptive GA for tandem-MS peptide sequencing (2006): Adaptive genetic algorithm for protein sequencing from tandem MS. Represents peptides as GA individuals, evolves them against the observed spectrum; the adaptive component tunes mutation / crossover rates on the fly.
  • ENPC (2006): Exhaustive composition enumeration
  • 157 nm photodissociation de novo sequencing (2010): De novo sequencing algorithm written for 157 nm photodissociation spectra, which yield clean x-ion series rather than the mixed b/y ions of CID. Reports per-residue accuracy and distinguishes leucine from isoleucine, which equal-mass residues normally make impossible.
  • LysNDeNovo (2010): Heuristic de novo sequencing for Lys-N peptides fragmented by ETD, exploiting their long c-ion series.
  • Positive charge derivatization de novo sequencing (2013): Positive charge derivatization + spectra merging
  • CycloAnt (2017): Hybrid ant colony optimisation with local search that reconstructs cyclic peptide sequences from distorted mass spectra.
  • Robust FL-Sequencing (2021): Full-length protein sequencing

Papers describing them (31)

Authors (92)

Albert J. R. Heck, Alexandros Georgiou, Alfred L. Yergey, Andrew N. Krutchinsky, Bas van Breukelen, Brian C. Bohrer, Brian T. Chait, Christian Rolando, D. Gostick, Dacheng He, E. Kapp, El-Ghazali Talbi, Frank Kjeldsen, Gabriel Padrón, Glen A. Humphrey, Gong Zhang, Gui‐Hua Yue, H. Matsuda, H. Wollnik, Haixu Tang, Hao‐Wei Song, Hong-Yan Shen, Hongying Zhong, Hong‐Hai Wang, I. Katakuse, J. Alex Taylor, J. Cottrell, J. Hoyes, J. Langridge, J. Skilling, J. Smith, James P. Reilly, James S. McElvain, Jean-Charles Boisson, Jennifer M. Campbell, Jens R. Coorssen, Jianguo Ji, Jing Wu, Jonathan A. Epstein, Jorge Fernández-de-Cossío, Joshua Zimmerberg, Klaus Biemann, Laetitia Jourdan, Li-Ming Xu, Liang Li, Liangyi Zhang, Luis Javier González, Lázaro Betancourt, Madalina M. Drugan, Mahmood A Rashid, Marvin L. Vestal, Matthew T. Olson, Michael L. Nielsen, Micromass UK Ltd, Mikhail M. Savitski, Mingrui An, Nadia Taouatas, Nobuaki Okumura, Paul S. Blank, Peng‐Yuan Yang, Peter S. Backlund, Peter Walden, Predrag Radivojac, Qing-Yu He, Qingsong Wang, R. Bordoli, R. Rachubinski, R. Wozniak, Randy J. Arnold, Richard S. Johnson, Rodion Demine, Roman A. Zubarev, Shabaz Mohammed, Sujata Baral, Swakkhar Shatabda, T. Matsuo, T. Sakurai, Takaki Shima, Toshifumi Takao, Vladimir Besada, Wenzhu Zhang, Xiao Zou, Xiaohui Liu, Xueyuan Xiao, Xuyang Zhao, Yasutsugu Shimonishi, Yong Fuga Li, Yoshinori Satomi, Yu Lu, Zhi-Biao Mai, Zhong-Hua Zhou, Zhongqi Zhang

Seen in the charts

Back to the full map

Back to top