The discovery of peptides possessing high biological activity is very challenging due to the enormous diversity for which only a minority have the desired properties.

To lower cost and reduce the time to obtain promising peptides , machine learning approaches can greatly assist in the process and even partly replace expensive laboratory experiments by learning a predictor with existing data or with a smaller amount of data generation.

Unfortunately, once the model is learned, selecting peptides having the greatest predicted bioactivity often requires a prohibitive amount of computational time.

Here, we focused on peptides as they have properties that make them excellent drug starting points.

Indeed, applying the model to every peptides would take an astronomical amount of computer time.

We demonstrate that this class of model has mathematical properties that makes it possible to rapidly identify and sort the best peptides .

For these reasons, this work will focus on using peptides as drug precursors.

However, it is important to note that combinatorial peptide chemistry cannot cover a significant part of the peptide diversity when peptides are longer than a few amino acids.

1) Phosphoproteo-mic data, like any mass spectrometry data, has missing values because many peptides are not analyzed by the detector, and using a “data not available” marker (NA) instead of zero facilitated calculation of statistical relationships based only on observed data [34].

For analysis of proteins, the total amount of phosphorylation of each protein was determined by summing peak intensity signals from all peptides for each protein in each sample.

In cases where conserved sequences did not allow unambiguous assignment to a particular protein, peptides were assigned to proteins that were detected by other phosphopeptides in the same sample or the first name was used.

1203 of these were tyrosine phosphorylated, identified from peptides immunoprecipitated using an anti-phosphotyrosine antibody.

Due to limits in mass spectrometric detection of peptides [43—47] , these data were not an exhaustive determination of all phosphorylated proteins in all samples.

In cases Where conserved sequences did not allow unambiguous assignment to a particular protein, the peptide name either retained multiple names, for example “FYN 420; LCK 394; SRC 419; YESl 426,” or were merged into all possible larger peptides , for example MAPKs and C-terminal inhibitory phosphorylations on SRC, FYN, and YESl (referred to as inclusively summed).

To identify patterns in tyrosine phosphorylation in neuroblastoma, we analyzed tyrosine phos-phoproteomic data acquired from 21 neuroblastoma cell lines using immunoprecipitation of tyrosine phosphorylated peptides as previously described [41,42].

Peak intensity was summed for each protein in each sample (i.e., cell line) using functions written in R [34], except in the case of SRC-family kinases (SFKs), where peptides phosphorylated on C-terminal inhibitory sites were tracked separately (denoted FYN_i, LYN_i, SRC_i, YES1_i, FRK_i).

Due to limits in mass spectrometry detection, data were not expected to be complete; for example SMS-KCN cells express NTRK2 (TrkB), but NTRK2 peptides were masked; and NTRK1 was not always detected in cell lines known to express it.

Peptides were inclusively summed for this phosphorylation site network (see Materials and Methods), rather than exclusively summed for the total phosphorylation protein network (81 Fig).

Phosphorylation sites represent the sum of all peptides surrounding that site; peptides whose conserved sequence is present in several proteins are indicated with multiple names, e.g., “FYN 420; LCK 394; SRC 419; YES1 426.” Fold changes are graphed on a blue-yellow color scale with blue representing a decrease, and yellow, an increase, compared to control (key).

Antimicrobial peptides are small, cationic proteins that can induce lysis of bacterial cells through interaction with their membranes.

Antimicrobial peptides have the ability to kill harmful bacteria through interaction With bacterial membranes.

While antimicrobial peptides have been the topic of many simulation studies, these studies have not incorporated the biochemical heterogeneity of natural membranes.

The peptides insert readily into the inner membrane, Whereas the interaction With the LPS-containing outer membrane is more complex.

Antimicrobial peptides (AMPS) are small cationic membrane-active peptides ; they can be found in most living organisms and play an essential part in innate immunity [1—3].

The peptides are thought to fulfil the initial stages of their bactericidal activity by anchoring themselves to the bacterial membrane via the DAB amino acids[12, 13].

Three of these additional PMB1 peptides became bound to the membrane during this period, resulting in a nine-PMB1-bound system, which was simulated for an additional 0.5 picrosecond (Sim_OM9).

In each of the simulations, the first peptides were observed to interact with the LPS sugars within only a few nanoseconds, where interaction is defined as atoms £0.35 nm apart, (Fig 1).

As demonstrated by the sudden drop in solvent accessible surface area, and by monitoring the mean centre of mass coordinate of the peptides along the Z-aXis (i.e.